Login
- Released At: 13-05-2026
- Page Views:
- Downloads:
- Table of Contents
- Related Documents
-
AD-NET Solution
Hardware Configuration Guide
Document version: 5W107-20250916
Copyright © 2026 New H3C Technologies Co., Ltd. All rights reserved.
No part of this manual may be reproduced or transmitted in any form or by any means without prior written consent of New H3C Technologies Co., Ltd.
Except for the trademarks of New H3C Technologies Co., Ltd., any trademarks that may be mentioned in this document are the property of their respective owners.
The information in this document is subject to change without notice.
Contents
Hardware requirements for AD-Campus
Deployment on physical servers (x86-64: Intel64/AMD64)
Standalone deployment of controller
Standalone deployment of SeerAnalyzer
Converged deployment of the controller and analyzer
Deployment on physical servers (x86-64 Hygon servers)
Standalone deployment of controller
Standalone deployment of SeerAnalyzer
Converged deployment of the controller and analyzer
Deployment on physical servers (ARM Kunpeng/Phytium servers)
Standalone deployment of controller
Standalone deployment of SeerAnalyzer
Converged deployment of the controller and analyzer
Hardware requirements for deployment on VMs
Requirements for test and demo deployment
Hardware requirements for AD-WAN carrier network deployment
Deployment on physical servers (x86-64(Intel64/AMD64))
Carrier device simulation analyzer deployment
Optional component deployment (x86-64(Intel64/AMD64))
Optional component deployment (x86-64 Hygon server/ARM Kunpeng server/ARM Phytium server)
Hardware requirements for AD-WAN branch scenarios
Hardware requirements for AD-WAN branch access network deployment (controller E6804 and later)
Deployment on physical servers
Hardware requirements for deployment on VMs
Hardware requirements for AD-WAN branch access network deployment (controller E6803 or earlier)
Hardware requirements for license server deployment
Overview
Introduction
This document provides hardware requirements for AD-NET products, including hardware requirements for single-controller deployment and converged deployment of controller and analyzer,
· For hardware requirements of AD-Campus scenarios, see "Hardware requirements for AD-Campus."
· For hardware requirements of AD-WAN scenarios, see "Hardware requirements for AD-WAN carrier network deployment."
· For hardware requirements of the license server, see "Hardware requirements for license server deployment."
Hardware requirements for AD-Campus
AD-Campus can be deployed in single-node mode or cluster mode, and can be deployed on a physical server or VM.
You can deploy the controller and the analyzer separately or in a converged manner.
If you deploy only SeerAnalyzer or deploy the controller and SeerAnalyzer separately in a cluster, the SeerAnalyzer nodes use load balancing mode. Decrease of SeerAnalyzer nodes will decrease analytics performance.
|
|
IMPORTANT: · For Unified Platform versions earlier than E0706 (including E06xx), the etcd partition requires a separate physical drive. · For Unified Platform E0706 and later versions, the etcd partition can share a physical drive with other partitions. As a best practice, use a separate physical drive for the etcd partition. |
Deployment on physical servers (x86-64: Intel64/AMD64)
Table 1 General requirements
|
Item |
Requirements |
|
|
Drive: · The drives must be configured in RAID 1, 5, or 10. · Storage controller: 1 GB cache, powerfail safeguard module supported, supercapacitor present. · The capacity requirement refers to the capacity after RAID setup. · 4k block file read/write IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. |
System drive |
SSDs or SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 2.4 TB. For HDDs, the rotation speed must be 7.2K RPM or higher. |
|
ETCD drive |
SSDs or 7.2K RPM or above SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 50 GB. Installation path: /var/lib/etcd. |
|
|
Data drive |
SSDs or SATA/SAS HDDs. As a best practice, configure RAID 5 by using three or more data drives. |
|
|
Network ports |
· Non-bonding mode: One NIC that has one 1Gbps network port. If the node also acts as an analyzer, one more network port is required. As a best practice, use a 10 Gbps network port. · Bonding mode (mode 2 and mode 4 are recommended): Two NICs, each NIC having two 1Gbps network ports. If the node also acts as an analyzer, two more network ports are required. As a best practice, use two 10 Gbps network ports to set up a Linux bonding interface. Upon converged deployment of the analyzer, controller, and Unified Platform, configure the controller and Unified Platform to share a network port, and configure the analyzer to use a separate northbound network port. If only one northbound network port is available, configure the controller and the northbound network of the analyzer to share a network port, and configure Unified Platform to use a separate network port. If the analyzer uses a unified northbound and southbound network, configure the analyzer and controller to share a network port. |
|
|
|
NOTE: If the NIC of the server provides 10-GE and GE ports, configure Unified Platform and the analyzer to share the 10-GE port and the controller to use the GE port alone. For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software. |
As a best practice, deploy the network devices in this ratio: (switches + ACs): APs =1: 3.
Four ONUs are counted as one switch in the network. For example, if the maximum number of switches allowed in the network is 100 and the network already has 50 switches, the number of ONUs allowed in the network is 200, which is calculated as follows: (100-50) × 4 = 200.
Standalone deployment of controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 2 Standalone deployment of controller on a single node (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 24 cores, 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
1 |
· CPU: 24 cores, 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
Table 3 Standalone deployment of controller in a cluster (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 22 core, 2.0 GHz. · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 22 core, 2.0 GHz. · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 24 cores, 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 24 cores, 2.0 GHz · Memory: 144 GB. · System drive: 2.7 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 26 cores, 2.0 GHz · Memory: 144 GB. · System drive: 3.0 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 36 cores, 2.0 GHz · Memory: 160 GB. · System drive: 3.2 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 36 cores, 2.0 GHz · Memory: 176 GB. · System drive: 3.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
Standalone deployment of SeerAnalyzer
This section describes only the hardware configuration for the required Unified Platform application packages and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the analyzer.
The hardware requirements for analyzer deployment vary by analyzer version. This document describes only the hardware requirements for SeerAnalyzer E62xx and later. For hardware requirements of SeerAnalyzer history versions, see "Hardware requirements for SeerAnalyzer history versions."
Table 4 Independent deployment of the analyzer in single-node mode (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
1 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 24 cores (total physical cores), 2.0 GHz · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 28 cores (total physical cores), 2.0 GHz · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 32 cores (total physical cores), 2.0 GHz · Memory: 288 GB · System drive: 3 TB after RAID setup · Data drive: 7 TB after RAID setup. 5 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 40 cores (total physical cores), 2.0 GHz · Memory: 384 GB. · System drive: 3 TB after RAID setup · Data drive: 11 TB after RAID setup. 8 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Table 5 Standalone deployment of SeerAnalyzer in a cluster (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
3 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 20 cores (total physical cores), 2.0 GHz · Memory: 160 GB. · System drive: 3 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 24 cores (total physical cores), 2.0 GHz · Memory: 192 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 28 cores (total physical cores), 2.0 GHz · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 32 cores (total physical cores), 2.0 GHz · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 8 TB after RAID setup. 6 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Converged deployment of the controller and analyzer
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
Converged deployment of the controller and analyzer supports the following modes:
· Converged deployment of the controller and analyzer in single-node mode. For the hardware requirements, see Table 6.
· Converged deployment of the controller and analyzer in 3+N cluster mode. In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. For hardware requirements, see the hardware requirements for standalone controller deployment and standalone analyzer deployment.
|
|
NOTE: For converged deployment of the controller and analyzer in 3+N cluster mode, the Oasis services required by the analyzer will be deployed on the master nodes. Therefore, in addition to the hardware resources required by the controller, you must also reserve the hardware resources required by the Oasis services on the master nodes. For the hardware requirements of the Oasis service, see Table 22. |
· Converged deployment of the controller and analyzer in a three-node cluster mode. In this mode, deploy the controller and analyzer on three master nodes. For hardware requirements, see Table 7.
This section describes the hardware requirements for converged deployment of SeerAnalyzer E62xx or later and the controller.
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
1 |
· CPU: 38 cores, 2.0 GHz · Memory: 272 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
1 |
· CPU: 38 cores, 2.0 GHz · Memory: 304 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
3 |
· CPU: 36 cores, 2.0 GHz · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 36 cores, 2.0 GHz · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 40 cores, 2.0 GHz · Memory: 240 GB. · System drive: 3.2 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 10000 online users · 2000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 40 cores, 2.0 GHz · Memory: 256 GB. · System drive: 3.8 TB (after RAID setup) · Data drive: 3 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 44 cores, 2.0 GHz · Memory: 288 GB. · System drive: 4.1 TB (after RAID setup) · Data drive: 4 TB or above after RAID setup. If HDDs are used, three HDDs of the same model are required. · ETCD drive: 50 GB or above after RAID setup. |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 56 cores, 2.0 GHz · Memory: 336 GB. · System drive: 4.3 TB (after RAID setup) · Data drive: 5 TB or above after RAID setup. If HDDs are used, four HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 58 cores, 2.0 GHz · Memory: 400 GB. · System drive: 4.5 TB (after RAID setup) · Data drive: 8 TB or above after RAID setup. If HDDs are used, six HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
DHCP server deployment
The following deployment modes are available:
· H3C vDHCP Server—vDHCP Server can assign IP addresses to devices during onboarding of the devices and assign IP addresses to users and endpoints that access the network. You can use vDHCP Server to assign IP addresses to users if the number of online endpoints does not exceed 50000. vDHCP Server and Unified Platform are installed on the same server or group of servers.
Table 8 Hardware requirements for vDHCP deployment in single-node or cluster mode
|
Node configuration |
Allocatable IP addresses |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
vDHCP node |
1 or 3 |
· CPU: 1 core, 2.0 GHz. · Memory: 2 GB. |
· ≤ 15000 |
|
vDHCP node |
1 or 3 |
· CPU: 1 core, 2.0 GHz. · Memory: 3 GB. |
· ≤ 50000 |
· Windows DHCP Server—If you do not use vDHCP Server to assign IP addresses to endpoints and no other DHCP server is deployed, you can use the DHCP server service provided with the Windows operating system to assign IP addresses to endpoints. Make sure the operating system is Windows Server 2012 R2 or a higher version. As a best practice, use Windows Server 2016. If the number of online endpoints is greater than 15000, the Windows DHCP Server must be deployed separately. As a best practice, deploy two Windows DHCP servers for HA.
Table 9 Hardware requirements for Windows DHCP server deployed on a physical server
|
Allocatable IP addresses |
Independent Windows DHCP server |
|
≤ 10000 |
· CPU: 8 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
10000 to 40000 |
· CPU: 16 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
40,000 to 60,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 32 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
60,000 to 100,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 64 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
|
IMPORTANT: As a best practice, configure the fail-permit DHCP server for the AD-Campus solution unless declared clearly by the user that is not required. As a best practice, uses a Windows DHCP server deployed in standalone mode as the fail-permit DHCP server. |
· Other third-party DHCP Server (for example, InfoBlox, BlueCat, and WRD)—User supplied. It cannot be connected to the AD-Campus controller, and it does support name-IP binding.
EPS deployment
EPS supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EPS. The hardware requirements for converged deployment of Unified Platform, controller, and EPS are the sum of the hardware requirements for them.
Table 10 Hardware requirements of EPS single-node deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 12 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
1 |
· CPU: 6 core, 2.0 GHz. · Memory: 16 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
Table 11 Hardware requirements for EPS cluster deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 6 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
|
EPS nodes |
3 |
· CPU: 6 core, 2.0 GHz. · Memory: 12 GB. · System drive: 300 GB after RAID setup |
≤ 50000 |
|
EPS nodes |
3 |
· CPU: 8 core, 2.0 GHz. · Memory: 16 GB. · System drive: 500 GB after RAID setup |
≤ 100000 |
SMP deployment
In the AD-Campus solution, unified network security management is supported and SMP can be installed based on Unified Platform. In converged deployment, security logging is not supported. The following tables describe hardware requirements only for SMP. The hardware requirements for converged deployment of Unified Platform, controller, and SMP are the sum of the hardware requirements for them.
In cluster deployment mode, SMP uses only the hardware resources on the master node.
Table 12 Hardware requirements for deploying SMP on physical servers
|
Security devices |
Requirements |
|
1 to 10 |
· CPU: 4 core, 2.0 GHz. · Memory: 10 GB. · Drive: 100 GB |
|
10 to 50 |
· CPU: 8 core, 2.0 GHz. · Memory: 16 GB. · Drive: 200 GB |
|
50 to 100 |
· CPU: 16 core, 2.0 GHz. · Memory: 32 GB. · Drive: 500 GB |
EIA deployment
EIA supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EIA. The hardware requirements for converged deployment of Unified Platform, controller, and EIA are the sum of the hardware requirements for them.
Table 13 Hardware requirements of EIA single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 24 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 24 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 28 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 28 GB. · System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 28 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
1 |
· CPU: 8 core, 2.0 GHz. · Memory: 32 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
1 |
· CPU: 8 core, 2.0 GHz. · Memory: 32 GB. · System drive: 500 GB after RAID setup |
100000 online users |
Table 14 Hardware requirements for EIA cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 16 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 20 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 20 GB. · System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 20 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
3 |
· CPU: 8 core, 2.0 GHz. · Memory: 24 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
3 |
· CPU: 8 core, 2.0 GHz. · Memory: 24 GB. · System drive: 500 GB after RAID setup |
100000 online users |
|
|
IMPORTANT: GlusterFS requires 50 GB more space. |
EAD deployment
EAD supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EAD. The hardware requirements for converged deployment of Unified Platform, controller, EIA, and EAD are the sum of the hardware requirements for them.
Table 15 Hardware requirements for EAD single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 2 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 2 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
10000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
20000 online users |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 18 GB. · System drive: 200 GB (after RAID setup) |
40000 online users |
Table 16 Hardware requirements for EAD cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 2 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
2000 online users |
|
Controller node |
3 |
· CPU: 2 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
5000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
10000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
20000 online users |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 16 GB. · System drive: 200 GB or above after RAID setup. |
40000 online users |
|
Controller node |
3 |
· CPU: 6 cores, 2.0 GHz · Memory: 24 GB. · System drive: 500 GB or above after RAID setup. |
100000 online users |
WSM deployment
Table 17 Software package description
|
Software package name |
Feature description |
Remarks |
|
WSM |
Basic WLAN management, including wireless device monitoring and configuration. |
Required. |
|
Oasis |
Intelligent WLAN analysis, including AP statistics, endpoint statistics, issue analysis, one-click diagnosis, one-click optimization, wireless security, Doctor AP, issue resolving, gradual optimization, and application analysis. |
Optional |
WSM supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for WSM. The hardware requirements for converged deployment of Unified Platform, controller, and WSM are the sum of the hardware requirements for them.
Table 18 Hardware requirements for WSM deployment in single-node mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 4 core, 2.0 GHz. · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
Table 19 Hardware requirements for WSM deployment in cluster mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 4 core, 2.0 GHz. · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 6 core, 2.0 GHz. · Memory: 16 GB. · System drive: 600 GB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 6 core, 2.0 GHz. · Memory: 20 GB. · System drive: 800 GB after RAID setup |
20000 online APs |
Table 20 Hardware requirements for WSM deployment in single-node mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 8 core, 2.0 GHz. · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 8 core, 2.0 GHz. · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
Table 21 Hardware requirements for WSM deployment in cluster mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 8 core, 2.0 GHz. · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 8 core, 2.0 GHz. · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 12 core, 2.0 GHz. · Memory: 40 GB. · System drive: 1 TB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 12 core, 2.0 GHz. · Memory: 52 GB. · System drive: 1.5 TB after RAID setup |
20000 online APs |
Oasis deployment
The following tables describe hardware requirements only for Oasis. The hardware requirements for converged deployment of Unified Platform, controller, and Oasis are the sum of the hardware requirements for them.
|
Node name |
Minimum single-node requirements |
Maximum resources that can be managed |
|
Controller node |
· CPU: 4 core, 2.0 GHz. · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online APs |
|
Controller node |
· CPU: 6 core, 2.0 GHz. · Memory: 24 GB. · System drive: 400 GB after RAID setup |
10000 online APs |
|
Controller node |
· CPU: 6 core, 2.0 GHz. · Memory: 32 GB. · System drive: 900 GB after RAID setup |
20000 online APs |
Deployment on physical servers (x86-64 Hygon servers)
|
|
IMPORTANT: For deployment on domestic servers, purchase domestic commercial operating systems as needed. |
Table 23 General requirements
|
Item |
Requirements |
|
|
Drive: · The drives must be configured in RAID 1, 5, or 10. · Storage controller: 1 GB cache, powerfail safeguard module supported, supercapacitor present. · The capacity requirement refers to the capacity after RAID setup. · 4k block file read/write IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. |
System drive |
SSDs or SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 2.4 TB. For HDDs, the rotation speed must be 7.2K RPM or higher. |
|
ETCD drive |
SSDs or 7.2K RPM or above SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 50 GB. Installation path: /var/lib/etcd. |
|
|
Data drive |
SSDs or SATA/SAS HDDs. As a best practice, configure RAID 5 by using three or more data drives. |
|
|
Network ports: |
· Non-bonding mode: One NIC that has one 1Gbps network port. If the node also acts as an analyzer, one more network port is required. As a best practice, use a 10 Gbps network port. · Bonding mode (mode 2 and mode 4 are recommended): Two NICs, each NIC having two 1Gbps network ports. If the node also acts as an analyzer, two more network ports are required. As a best practice, use two 10 Gbps network ports to set up a Linux bonding interface. Upon converged deployment of the analyzer, controller, and Unified Platform, configure the controller and Unified Platform to share a network port, and configure the analyzer to use a separate northbound network port. If only one northbound network port is available, configure the controller and the northbound network of the analyzer to share a network port, and configure Unified Platform to use a separate network port. If the analyzer uses a unified northbound and southbound network, configure the analyzer and controller to share a network port. |
|
|
CPU |
Hygon C86 7265, 2.2 GHz |
|
|
|
NOTE: If the NIC of the server provides 10-GE and GE ports, configure Unified Platform and the analyzer to share the 10-GE port and the controller to use the GE port alone. For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software. |
As a best practice, deploy the network devices in this ratio: (switches + ACs): APs =1: 3.
Four ONUs are counted as one switch in the network. For example, if the maximum number of switches allowed in the network is 100 and the network already has 50 switches, the number of ONUs allowed in the network is 200, which is calculated as follows: (100-50) × 4 = 200.
Standalone deployment of controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 24 Standalone deployment of controller on a single node (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
1 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
Table 25 Standalone deployment of controller in a cluster (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 32 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 32 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 40 cores · Memory: 144 GB. · System drive: 2.7 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 42 cores · Memory: 144 GB. · System drive: 3.0 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 54 cores · Memory: 176 GB. · System drive: 3.2 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 58 cores · Memory: 192 GB. · System drive: 3.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
Standalone deployment of SeerAnalyzer
This section describes only the hardware configuration for the required Unified Platform application packages and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the analyzer.
The hardware requirements for analyzer deployment vary by analyzer version. This document describes only the hardware requirements for SeerAnalyzer E62xx and later. For hardware requirements of SeerAnalyzer history versions, see "Hardware requirements for SeerAnalyzer history versions."
Table 26 Independent deployment of the analyzer in single-node mode (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 36 cores · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 42 cores · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 48 cores · Memory: 288 GB · System drive: 3 TB after RAID setup · Data drive: 7 TB after RAID setup. 5 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 60 cores · Memory: 384 GB. · System drive: 3 TB after RAID setup · Data drive: 11 TB after RAID setup. 8 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Table 27 Standalone deployment of SeerAnalyzer in a cluster (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 160 GB. · System drive: 3 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 36 cores · Memory: 192 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 42 cores · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 48 cores · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 8 TB after RAID setup. 6 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Converged deployment of the controller and analyzer
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
Converged deployment of the controller and analyzer supports the following modes:
· Converged deployment of the controller and analyzer in single-node mode. For the hardware requirements, see Table 6.
· Converged deployment of the controller and analyzer in 3+N cluster mode. In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. For hardware requirements, see the hardware requirements for standalone controller deployment and standalone analyzer deployment.
|
|
NOTE: For converged deployment of the controller and analyzer in 3+N cluster mode, the Oasis services required by the analyzer will be deployed on the master nodes. Therefore, in addition to the hardware resources required by the controller, you must also reserve the hardware resources required by the Oasis services on the master nodes. For the hardware requirements of the Oasis service, see Table 43. |
· Converged deployment of the controller and analyzer in a three-node cluster mode. In this mode, deploy the controller and analyzer on three master nodes. For hardware requirements, see Table 7.
This section describes the hardware requirements for converged deployment of SeerAnalyzer E62xx or later and the controller.
Table 28 Converged deployment of the controller and analyzer in single-node mode (Unified Platform + vDHCP + SE + EIA + WSM + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
1 |
· CPU: 56 cores · Memory: 288 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
1 |
· CPU: 56 cores · Memory: 320 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
Table 29 Converged deployment of the controller and analyzer in three-node cluster mode (Unified Platform + vDHCP + SE + EIA + WSM + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
3 |
· CPU: 54 cores · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 54 cores · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 60 cores · Memory: 256 GB. · System drive: 3.2 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 10000 online users · 2000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 60 cores · Memory: 266 GB. · System drive: 3.8 TB (after RAID setup) · Data drive: 3 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 70 cores · Memory: 288 GB. · System drive: 4.1 TB (after RAID setup) · Data drive: 4 TB or above after RAID setup. If HDDs are used, three HDDs of the same model are required. · ETCD drive: 50 GB or above after RAID setup. |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 84 cores · Memory: 352 GB. · System drive: 4.3 TB (after RAID setup) · Data drive: 5 TB or above after RAID setup. If HDDs are used, four HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 90 cores · Memory: 416 GB. · System drive: 4.5 TB (after RAID setup) · Data drive: 8 TB or above after RAID setup. If HDDs are used, six HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
DHCP server deployment
The following deployment modes are available:
· UNIS vDHCP Server—vDHCP Server can assign IP addresses to devices during onboarding of the devices and assign IP addresses to users and endpoints that access the network. You can use vDHCP Server to assign IP addresses to users if the number of online endpoints does not exceed 50000. vDHCP Server and Unified Platform are installed on the same server or group of servers.
Table 30 Hardware requirements for vDHCP deployment in single-node or cluster mode
|
Node configuration |
Allocatable IP addresses |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
vDHCP node |
1 or 3 |
· CPU: 1 cores · Memory: 2 GB. |
≤ 15000 |
|
vDHCP node |
1 or 3 |
· CPU: 1 cores · Memory: 3 GB. |
≤ 50000 |
· Windows DHCP Server—If you do not use vDHCP Server to assign IP addresses to endpoints and no other DHCP server is deployed, you can use the DHCP server service provided with the Windows operating system to assign IP addresses to endpoints. Make sure the operating system is Windows Server 2012 R2 or a higher version. As a best practice, use Windows Server 2016. If the number of online endpoints is greater than 15000, the Windows DHCP Server must be deployed separately. As a best practice, deploy two Windows DHCP servers for HA.
Table 31 Hardware requirements for Windows DHCP server deployed on a physical server
|
Allocatable IP addresses |
Independent Windows DHCP server |
|
≤ 10000 |
· CPU: 8 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
10000 to 40000 |
· CPU: 16 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
40,000 to 60,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 32 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
60,000 to 100,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 64 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
|
IMPORTANT: As a best practice, configure the fail-permit DHCP server for the SDN-Campus solution unless declared clearly by the user that is not required. As a best practice, uses a Windows DHCP server deployed in standalone mode as the fail-permit DHCP server. |
· Other third-party DHCP Server (for example, InfoBlox, BlueCat, and WRD)—User supplied. It cannot be connected to the SDN-Campus controller, and it does support name-IP binding.
EPS deployment
EPS supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EPS. The hardware requirements for converged deployment of Unified Platform, controller, and EPS are the sum of the hardware requirements for them.
Table 32 Hardware requirements of EPS single-node deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
1 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
1 |
· CPU: 10 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
Table 33 Hardware requirements for EPS cluster deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
3 |
· CPU: 6 cores · Memory: 6 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
3 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
|
EPS nodes |
3 |
· CPU: 10 cores · Memory: 12 GB. · System drive: 300 GB after RAID setup |
≤ 50000 |
|
EPS nodes |
3 |
· CPU: 12 cores · Memory: 16 GB. · System drive: 500 GB after RAID setup |
≤ 100000 |
EIA deployment
EIA supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EIA. The hardware requirements for converged deployment of Unified Platform, controller, and EIA are the sum of the hardware requirements for them.
Table 34 Hardware requirements of EIA single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 6 cores · Memory: 24 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 24 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 32 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 32 GB. · System drive: 500 GB after RAID setup |
100000 online users |
Table 35 Hardware requirements for EIA cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. · System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 500 GB after RAID setup |
100000 online users |
|
|
IMPORTANT: GlusterFS requires 50 GB more space. |
EAD deployment
EAD supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EAD. The hardware requirements for converged deployment of Unified Platform, controller, EIA, and EAD are the sum of the hardware requirements for them.
Table 36 Hardware requirements for EAD single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
10000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
20000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 18 GB. · System drive: 200 GB (after RAID setup) |
40000 online users |
Table 37 Hardware requirements for EAD cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
2000 online users |
|
Controller node |
3 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
5000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
10000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
20000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB or above after RAID setup. |
40000 online users |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 24 GB. · System drive: 500 GB or above after RAID setup. |
100000 online users |
WSM deployment
Table 38 Software package description
|
Software package name |
Feature description |
Remarks |
|
WSM |
Basic WLAN management, including wireless device monitoring and configuration. |
Required. |
|
Oasis |
Intelligent WLAN analysis, including AP statistics, endpoint statistics, issue analysis, one-click diagnosis, one-click optimization, wireless security, Doctor AP, issue resolving, gradual optimization, and application analysis. |
Optional |
WSM supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for WSM. The hardware requirements for converged deployment of Unified Platform, controller, and WSM are the sum of the hardware requirements for them.
Table 39 Hardware requirements for WSM deployment in single-node mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
Table 40 Hardware requirements for WSM deployment in cluster mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 16 GB. · System drive: 600 GB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 20 GB. · System drive: 800 GB after RAID setup |
20000 online APs |
Table 41 Hardware requirements for WSM deployment in single-node mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
Table 42 Hardware requirements for WSM deployment in cluster mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 18 cores · Memory: 40 GB. · System drive: 1 TB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 18 cores · Memory: 52 GB. · System drive: 1.5 TB after RAID setup |
20000 online APs |
Oasis deployment
The following tables describe hardware requirements only for Oasis. The hardware requirements for converged deployment of Unified Platform, controller, and Oasis are the sum of the hardware requirements for them.
|
Node name |
Minimum single-node requirements |
Maximum resources that can be managed |
|
Controller node |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online APs |
|
Controller node |
· CPU: 9 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
10000 online APs |
|
Controller node |
· CPU: 9 cores · Memory: 32 GB. · System drive: 900 GB after RAID setup |
20000 online APs |
Deployment on physical servers (ARM Kunpeng/Phytium servers)
|
|
IMPORTANT: For deployment on domestic servers, purchase domestic commercial operating systems as needed. |
Table 44 General requirements
|
Item |
Requirements |
|
|
Drive: · The drives must be configured in RAID 1, 5, or 10. · Storage controller: 1 GB cache, powerfail safeguard module supported, supercapacitor present. · The capacity requirement refers to the capacity after RAID setup. · 4k block file read/write IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. |
System drive |
SSDs or SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 2.4 TB. For HDDs, the rotation speed must be 7.2K RPM or higher. |
|
ETCD drive |
SSDs or 7.2K RPM or above SATA/SAS HDDs configured in RAID that provides a minimum total drive size of 50 GB. Installation path: /var/lib/etcd. |
|
|
Data drive |
SSDs or SATA/SAS HDDs. As a best practice, configure RAID 5 by using three or more data drives. |
|
|
Network ports: |
· Non-bonding mode: One NIC that has one 1Gbps network port. If the node also acts as an analyzer, one more network port is required. As a best practice, use a 10 Gbps network port. · Bonding mode (mode 2 and mode 4 are recommended): Two NICs, each NIC having two 1Gbps network ports. If the node also acts as an analyzer, two more network ports are required. As a best practice, use two 10 Gbps network ports to set up a Linux bonding interface. Upon converged deployment of the analyzer, controller, and Unified Platform, configure the controller and Unified Platform to share a network port, and configure the analyzer to use a separate northbound network port. If only one northbound network port is available, configure the controller and the northbound network of the analyzer to share a network port, and configure Unified Platform to use a separate network port. If the analyzer uses a unified northbound and southbound network, configure the analyzer and controller to share a network port. |
|
|
CPU |
· Kunpeng 920, 2.6GHz · Phytium S2500, 2.1 GHz |
|
|
|
NOTE: If the NIC of the server provides 10-GE and GE ports, configure Unified Platform and the analyzer to share the 10-GE port and the controller to use the GE port alone. For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software. |
As a best practice, deploy the network devices in this ratio: (switches + ACs): APs =1: 3.
Four ONUs are counted as one switch in the network. For example, if the maximum number of switches allowed in the network is 100 and the network already has 50 switches, the number of ONUs allowed in the network is 200, which is calculated as follows: (100-50) × 4 = 200.
Standalone deployment of controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 45 Standalone deployment of controller on a single node (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
1 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
Table 46 Standalone deployment of controller in a cluster (Unified Platform + vDHCP + SE + EIA + WSM (Oasis excluded) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 32 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 32 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 36 cores · Memory: 144 GB. · System drive: 2.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Controller node |
3 |
· CPU: 40 cores · Memory: 144 GB. · System drive: 2.7 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 42 cores · Memory: 144 GB. · System drive: 3.0 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 54 cores · Memory: 176 GB. · System drive: 3.2 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller node |
3 |
· CPU: 58 cores · Memory: 192 GB. · System drive: 3.4 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
Standalone deployment of SeerAnalyzer
This section describes only the hardware configuration for the required Unified Platform application packages and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the analyzer.
The analyzer cannot be deployed on Phytium servers.
Table 47 Independent deployment of the analyzer in single-node mode (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 30 cores · Memory: 192 GB. · System drive: 2.4 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 36 cores · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 42 cores · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 48 cores · Memory: 288 GB · System drive: 3 TB after RAID setup · Data drive: 7 TB after RAID setup. 5 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
1 |
· CPU: 60 cores · Memory: 384 GB. · System drive: 3 TB after RAID setup · Data drive: 11 TB after RAID setup. 8 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Table 48 Standalone deployment of SeerAnalyzer in a cluster (Unified Platform + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 2000 online users · 400 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 5000 online users · 1000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 128 GB. · System drive: 2.4 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 10000 online users · 2000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 30 cores · Memory: 160 GB. · System drive: 3 TB after RAID setup · Data drive: 3 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 20000 online users · 4000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 36 cores · Memory: 192 GB. · System drive: 3 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 40000 online users · 8000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 42 cores · Memory: 224 GB. · System drive: 3 TB after RAID setup · Data drive: 5 TB after RAID setup. 4 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 60000 online users · 12000 switches, ACs, and APs in total |
|
Analyzer node |
3 |
· CPU: 48 cores · Memory: 256GB · System drive: 3 TB after RAID setup · Data drive: 8 TB after RAID setup. 6 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. |
· 100000 online users · 20000 switches, ACs, and APs in total |
Converged deployment of the controller and analyzer
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
Converged deployment of the controller and analyzer supports the following modes:
· Converged deployment of the controller and analyzer in single-node mode. For the hardware requirements, see Table 6.
· Converged deployment of the controller and analyzer in 3+N cluster mode. In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. For hardware requirements, see the hardware requirements for standalone controller deployment and standalone analyzer deployment.
|
|
NOTE: For converged deployment of the controller and analyzer in 3+N cluster mode, the Oasis services required by the analyzer will be deployed on the master nodes. Therefore, in addition to the hardware resources required by the controller, you must also reserve the hardware resources required by the Oasis services on the master nodes. For the hardware requirements of the Oasis service, see Table 64. |
· Converged deployment of the controller and analyzer in a three-node cluster mode. In this mode, deploy the controller and analyzer on three master nodes. For hardware requirements, see Table 7.
This section describes the hardware requirements for converged deployment of SeerAnalyzer E62xx or later and the controller. The analyzer cannot be deployed on Phytium servers.
Table 49 Converged deployment of the controller and analyzer in single-node mode (Unified Platform + vDHCP + SE + EIA + WSM + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
1 |
· CPU: 56 cores · Memory: 288 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
1 |
· CPU: 56 cores · Memory: 320 GB. · System drive: 2.4 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
Table 50 Converged deployment of the controller and analyzer in three-node cluster mode (Unified Platform + vDHCP + SE + EIA + WSM + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
3 |
· CPU: 54 cores · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 2000 online users · 400 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 54 cores · Memory: 224 GB. · System drive: 3 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 5000 online users · 1000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 60 cores · Memory: 256 GB. · System drive: 3.2 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 10000 online users · 2000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 60 cores · Memory: 266 GB. · System drive: 3.8 TB (after RAID setup) · Data drive: 3 TB or above after RAID setup. If HDDs are used, two HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 20000 online users · 4000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 70 cores · Memory: 288 GB. · System drive: 4.1 TB (after RAID setup) · Data drive: 4 TB or above after RAID setup. If HDDs are used, three HDDs of the same model are required. · ETCD drive: 50 GB or above after RAID setup. |
· 40000 online users · 8000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 84 cores · Memory: 352 GB. · System drive: 4.3 TB (after RAID setup) · Data drive: 5 TB or above after RAID setup. If HDDs are used, four HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 60000 online users · 12000 devices, including switches, ACs, and APs. |
|
Controller and analyzer |
3 |
· CPU: 90 cores · Memory: 416 GB. · System drive: 4.5 TB (after RAID setup) · Data drive: 8 TB or above after RAID setup. If HDDs are used, six HDDs of the same type are required. · ETCD drive: 50 GB or above after RAID setup. |
· 100000 online users · 20000 devices, including switches, ACs, and APs. |
DHCP server deployment
The following deployment modes are available:
· UNIS vDHCP Server—vDHCP Server can assign IP addresses to devices during onboarding of the devices and assign IP addresses to users and endpoints that access the network. You can use vDHCP Server to assign IP addresses to users if the number of online endpoints does not exceed 50000. vDHCP Server and Unified Platform are installed on the same server or group of servers.
Table 51 Hardware requirements for vDHCP deployment in single-node or cluster mode
|
Node configuration |
Allocatable IP addresses |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
vDHCP node |
1 or 3 |
· CPU: 1 cores · Memory: 2 GB. |
≤ 15000 |
|
vDHCP node |
1 or 3 |
· CPU: 1 cores · Memory: 3 GB. |
≤ 50000 |
· Windows DHCP Server—If you do not use vDHCP Server to assign IP addresses to endpoints and no other DHCP server is deployed, you can use the DHCP server service provided with the Windows operating system to assign IP addresses to endpoints. Make sure the operating system is Windows Server 2012 R2 or a higher version. As a best practice, use Windows Server 2016. If the number of online endpoints is greater than 15000, the Windows DHCP Server must be deployed separately. As a best practice, deploy two Windows DHCP servers for HA.
Table 52 Hardware requirements for Windows DHCP server deployed on a physical server
|
Allocatable IP addresses |
Independent Windows DHCP server |
|
≤ 10000 |
· CPU: 8 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
10000 to 40000 |
· CPU: 16 cores, 2.0 GHz · Memory: 16 GB · Drive: 2 × 300 GB drives in RAID 1 · RAID controller: 256 MB cache |
|
40,000 to 60,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 32 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
60,000 to 100,000 |
· CPU: 24 cores, 2.0 GHz · Memory: 64 GB · Drive: 2 × 500 GB drives in RAID 1 · RAID controller: 1 GB cache |
|
|
IMPORTANT: As a best practice, configure the fail-permit DHCP server for the SDN-Campus solution unless declared clearly by the user that is not required. As a best practice, uses a Windows DHCP server deployed in standalone mode as the fail-permit DHCP server. |
· Other third-party DHCP Server (for example, InfoBlox, BlueCat, and WRD)—User supplied. It cannot be connected to the SDN-Campus controller, and it does support name-IP binding.
EPS deployment
EPS supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EPS. The hardware requirements for converged deployment of Unified Platform, controller, and EPS are the sum of the hardware requirements for them.
Table 53 Hardware requirements of EPS single-node deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
1 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
1 |
· CPU: 10 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
Table 54 Hardware requirements for EPS cluster deployment
|
Node configuration |
Maximum endpoints that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
EPS nodes |
3 |
· CPU: 6 cores · Memory: 6 GB. · System drive: 100 GB after RAID setup |
≤ 10000 |
|
EPS nodes |
3 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
≤ 20000 |
|
EPS nodes |
3 |
· CPU: 10 cores · Memory: 12 GB. · System drive: 300 GB after RAID setup |
≤ 50000 |
|
EPS nodes |
3 |
· CPU: 12 cores · Memory: 16 GB. · System drive: 500 GB after RAID setup |
≤ 100000 |
EIA deployment
EIA supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EIA. The hardware requirements for converged deployment of Unified Platform, controller, and EIA are the sum of the hardware requirements for them.
Table 55 Hardware requirements of EIA single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 6 cores · Memory: 24 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 24 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 28 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 32 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 32 GB. · System drive: 500 GB after RAID setup |
100000 online users |
Table 56 Hardware requirements for EIA cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. · System drive: 300 GB after RAID setup |
10000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. System drive: 300 GB after RAID setup |
20000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 20 GB. · System drive: 300 GB after RAID setup |
40000 online users |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 500 GB after RAID setup |
60000 online users |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 500 GB after RAID setup |
100000 online users |
|
|
IMPORTANT: GlusterFS requires 50 GB more space. |
EAD deployment
EAD supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for EAD. The hardware requirements for converged deployment of Unified Platform, controller, EIA, and EAD are the sum of the hardware requirements for them.
Table 57 Hardware requirements for EAD single-node deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online users |
|
Controller node |
1 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
5000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
10000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 14 GB. · System drive: 200 GB (after RAID setup) |
20000 online users |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 18 GB. · System drive: 200 GB (after RAID setup) |
40000 online users |
Table 58 Hardware requirements for EAD cluster deployment
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
2000 online users |
|
Controller node |
3 |
· CPU: 4 cores · Memory: 8 GB. · System drive: 200 GB or above after RAID setup. |
5000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
10000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 200 GB or above after RAID setup. |
20000 online users |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB or above after RAID setup. |
40000 online users |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 24 GB. · System drive: 500 GB or above after RAID setup. |
100000 online users |
WSM deployment
Table 59 Software package description
|
Software package name |
Feature description |
Remarks |
|
WSM |
Basic WLAN management, including wireless device monitoring and configuration. |
Required. |
|
Oasis |
Intelligent WLAN analysis, including AP statistics, endpoint statistics, issue analysis, one-click diagnosis, one-click optimization, wireless security, Doctor AP, issue resolving, gradual optimization, and application analysis. |
Optional |
WSM supports single-node deployment and cluster deployment. The following tables describe hardware requirements only for WSM. The hardware requirements for converged deployment of Unified Platform, controller, and WSM are the sum of the hardware requirements for them.
Table 60 Hardware requirements for WSM deployment in single-node mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
Table 61 Hardware requirements for WSM deployment in cluster mode (including only WSM)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 6 cores · Memory: 8 GB. · System drive: 200 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 6 cores · Memory: 12 GB. · System drive: 400 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 16 GB. · System drive: 600 GB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 10 cores · Memory: 20 GB. · System drive: 800 GB after RAID setup |
20000 online APs |
Table 62 Hardware requirements for WSM deployment in single-node mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
1 |
· CPU: 12 cores · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
Table 63 Hardware requirements for WSM deployment in cluster mode (including WSM and Oasis)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
3 |
· CPU: 12 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
2000 online APs |
|
Controller node |
3 |
· CPU: 12 cores · Memory: 28 GB. · System drive: 600 GB after RAID setup |
5000 online APs |
|
Controller node |
3 |
· CPU: 18 cores · Memory: 40 GB. · System drive: 1 TB after RAID setup |
10000 online APs |
|
Controller node |
3 |
· CPU: 18 cores · Memory: 52 GB. · System drive: 1.5 TB after RAID setup |
20000 online APs |
Oasis deployment
The following tables describe hardware requirements only for Oasis. The hardware requirements for converged deployment of Unified Platform, controller, and Oasis are the sum of the hardware requirements for them.
|
Node name |
Minimum single-node requirements |
Maximum resources that can be managed |
|
Controller node |
· CPU: 6 cores · Memory: 16 GB. · System drive: 200 GB after RAID setup |
5000 online APs |
|
Controller node |
· CPU: 9 cores · Memory: 24 GB. · System drive: 400 GB after RAID setup |
10000 online APs |
|
Controller node |
· CPU: 9 cores · Memory: 32 GB. · System drive: 900 GB after RAID setup |
20000 online APs |
Hardware requirements for deployment on VMs
The AD-Campus solution supports deployment of the controller and analyzer on VMs. The supported hypervisors and hypervisor versions are the same as those supported by Unified Platform.
For deployment on a VM, if the host where the VM resides is enabled with hyperthreading, the number of required vCPUs is twice the number of CPU cores required for deployment on a physical server. If the host is not enabled with hyperthreading, the number of required vCPUs is the same as the number of CPU cores required for deployment on a physical server. The memory and drive resources required for deployment on a VM are the same as those required for deployment on a physical server. As a best practice, enable hyperthreading.
For deployment on VMs, follow these restrictions and guidelines:
· VM deployment requires VMware ESXi 6.7.0, H3C_CAS-E0706, or their later versions.
· 4k block file read/write IOPS ≥ 10000. When shared storage is used, the IOPS performance degrades as the number of users increases. Make sure the IOPS performance meets the requirements.
· VM migration is not supported.
· Do not deploy the VMs in the cluster on the same host.
· Do not enable overcommitment on the VMs. To ensure system stability, reserve resources for the controller VM.
· Do not set the simplified mode for the VM storage volumes.
· The VM cluster nodes must be installed separately. You cannot clone a node and then modify its IP.
· As a best practice, install Unified Platform on a VM through mounting an ISO image rather than building a hard disk image.
· Do not deploy virtual cluster nodes and other IO-intensive VMs on the same server.
· When you specify an etcd partition, make sure the data disks do not use the same storage pool. If you have an SSD storage pool, as a best practice, use the SSD storage pool to allocate an etcd partition.
· To deploy the controller on a VMware VM, you must enable promiscuous mode and pseudo-transmission on the host where the VM resides.
|
|
IMPORTANT: For Unified Platform versions earlier than E0706 (including E06xx), the etcd partition requires a separate physical drive. For Unified Platform E0706 and later versions, the etcd partition can share a physical drive with other partitions. As a best practice, use a separate physical drive for the etcd partition. |
Requirements for test and demo deployment
The hardware configuration in this section is only used for test and demonstration, and cannot be used in site scenarios.
|
|
IMPORTANT: Performance test is not supported with this resource configuration |
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 12 cores, 2.0 GHz · Memory: 88 GB. · System drive: 2 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 100 online users · 20 switches |
Table 66 Hardware requirements of controller deployment for small-scale testing (Unified Platform + vDHCP + SE + EIA + EAD + EPS + WSM (including Oasis) )
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller node |
1 |
· CPU: 16 cores, 2.0 GHz · Memory: 104 GB. · System drive: 2 TB (after RAID setup) · ETCD drive: 50 GB (after RAID setup) |
· 100 online users · 20 switches, ACs, and APs in total |
Table 67 Hardware requirements of analyzer deployment for small-scale testing (Unified Platform + vDHCP + SE + EIA + WSM + SA)
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Controller and analyzer |
1 |
· CPU: 22 cores, 2.0 GHz · Memory: 224 GB. · System drive: 2 TB (after RAID setup) · Data drive: 2 TB or above after RAID setup · ETCD drive: 50 GB (after RAID setup) |
· 100 online users · 20 switches, ACs, and APs in total |
Hardware requirements for AD-WAN carrier network deployment
The AD-WAN carrier network supports single-node deployment and cluster deployment, and can be deployed on physical servers or VMs.
You can deploy the controller and the analyzer separately or in a converged manner.
When the controller and analyzer are deployed in a converged manner, you must deploy a DTN physical host if device simulation is used.
|
|
NOTE: · For Unified Platform versions earlier than E0706 (including E06xx), the etcd partition requires a separate physical drive. · For Unified Platform E0706 and later versions, the etcd partition can share a physical drive with other partitions. As a best practice, use a separate physical drive for the etcd partition. · The configuration in this section does not involve that for the network management services (features except for RES and PERF). To use complete network management services, add additional resources for them. |
Deployment on physical servers (x86-64(Intel64/AMD64))
General disk requirements
This section introduces the general disk configuration of physical servers, including recommended RAID mode, disk type, and performance requirements. For specific disk capacity requirements, see the configuration requirements in the corresponding scenarios.
Table 68 General requirements
|
Item |
Requirements |
||
|
General disk requirements: · Configure RAID 1, RAID 5, or RAID 10. · Use SSDs or 7.2K RPM or above HDDs. IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. · For HDDs, the RAID controller must have 1 GB cache memory, support power fail protection, and have a supercapacitor in place. |
System disk |
SSD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2 × 1920GB disks · RAID5: 3 × 960 GB or 5 × 480 GB disks · RAID10: 4 × 960 GB or 8 × 480 GB disks |
|
HDD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2 × 1920GB disks · RAID5: 3 × 1200 GB or 5 × 600 GB disks · RAID10: 4 × 1200 GB or 8 × 600 GB disks |
||
|
ETCD disk |
SSD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2 × 480GB disks |
|
|
HDD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2 × 600GB disks |
||
|
Data disk |
A minimum of 8 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of three disks of the same model A minimum of 12 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of five disks of the same model A minimum of 24 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of seven disks of the same model |
||
|
|
NOTE: For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software. |
Standalone deployment of the carrier controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 69 Carrier controller deployment in cluster mode (Unified Platform+controller)
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller |
200 |
Intel Xeon V3 series or later, total cores ≥ 12 Clock speed: 2.0 GHz or above |
96 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
· Non-bonding mode: 2 × 1 Gbps or above ports. · Bonding mode: 4 × 1 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller |
2000 |
Intel Xeon V3 series or later, total cores ≥ 16 Clock speed: 2.0 GHz or above |
128 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
|
Table 70 Carrier controller deployment in single-node mode (Unified Platform+controller)
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller |
200 |
Intel Xeon V3 series or later, total cores ≥ 14 Clock speed: 2.0 GHz or above |
96 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
· Non-bonding mode: 2 × 1 Gbps or above ports. · Bonding mode: 4 × 1 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller |
2000 |
Intel Xeon V3 series or later, total cores ≥ 18 Clock speed: 2.0 GHz or above |
144 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
|
Converged deployment of the AD-WAN carrier controller
This section describes the hardware requirements for converged deployment of SeerAnalyzer E62xx or later and the controller.
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
The following converged deployment modes are available:
· Converged deployment of the controller and analyzer in single-node mode. For the hardware requirements, see Table 97.
· Converged deployment of the controller and analyzer in a three-node cluster mode. In this mode, deploy the controller and analyzer on three master nodes. For hardware requirements, see Table 98.
· Converged deployment of the controller and analyzer in 3+N cluster mode. In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. Table 99 shows the hardware requirements for the carrier controller in this mode. Table 100 shows the hardware requirements for the analyzer in 3+1 cluster mode. Table 101 shows the hardware requirements for the analyzer in 3+3 cluster mode.
|
|
NOTE: In 3+N cluster deployment mode, the analyzer is deployed on worker nodes. In this scenario, the hardware requirements vary by analyzer version. This document describes only the hardware requirements for SeerAnalyzer E62xx and later. For hardware requirements of SeerAnalyzer history versions, see "Hardware requirements for SeerAnalyzer history versions." |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller + analyzer |
200 |
Intel Xeon V3 or higher, 30 or more cores 2.0 GHz or above |
208 GB or more
|
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 2 TB or above. Use 2 or more drives of the same model |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
|
|
IMPORTANT: In single-node deployment mode, follow these restrictions and guidelines: · The single-node deployment mode cannot provide redundancy, and device failures will cause service interruption. As a best practice, configure a remote backup server. · If the controller and analyzer are deployed in a converged manner on a single node, you cannot scale up the system to a cluster. · For analyzers in a version earlier than E6502, you must deploy a simulation analyzer when deploying the simulation feature. The simulation analyzer requires an exclusive Unified Platform. For more information, see H3C SeerAnalyzer-WAN Simulation Network Configuration Guide. |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller + analyzer |
200 |
Intel Xeon V3 or higher, 28 or more cores 2.0 GHz or above |
176 GB or more
|
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
|
Controller + analyzer |
2000 |
Intel Xeon V3 or higher, 32 or more cores 2.0 GHz or above |
320 GB or more
|
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 4 TB or above. Use 3 or more drives of the same model |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller |
200 |
Intel Xeon V3 or higher, 20 or more cores 2.0 GHz or above |
126 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 1 Gbps network ports, each two forming a Linux bonding interface. |
|
Controller |
2000 |
Intel Xeon V3 or higher, 24 or more cores 2.0 GHz or above |
174 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
1 |
· CPU: 24 cores (total physical cores), 2.0 GHz · Memory: 160 GB. · System drive: 1.92 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. · Network ports: ¡ Non-bonding mode: 2 × 10Gbps Ethernet ports. ¡ Bonding mode (recommended mode: mode 2 or mode 4): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
200 |
|
Node configuration |
Maximum number of devices |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
3 |
· CPU: 18 cores (total physical cores), 2.0 GHz · Memory: 140 GB. · System drive: 1.92 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd · Network ports: ¡ Non-bonding mode: 2 × 10Gbps Ethernet ports. ¡ Bonding mode (recommended mode: mode 2 or mode 4): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
2000 |
Arbitration server deployment
Table 76 Arbitration server deployment
|
Node name |
Node quantity |
Single-node requirements |
|
Arbitration node |
1 |
CPU: 2 cores, 2.0 GHz Memory: 16 GB Disk: Configure RAID1 or RAID10. Select one of the following configuration options: · Disk configuration 1: ¡ System disk: 2 × 480 GB SSDs in RAID1 (the total disk capacity after RAID setup must be greater than or equal to 256 GB), IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. ¡ ETCD disk: 2 × 480 GB SSDs in RAID1 (the total disk capacity after RAID setup must be greater than or equal to 50GB), installation path: /var/lib/etcd · Disk configuration 2: ¡ System disk: 2 × 600 GB 7.2K RPM or above HDDs in RAID1 (the total disk capacity after RAID setup must be greater than or equal to 256 GB), IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. ¡ ETCD disk: 2 × 600 GB 7.2K RPM or above HDDs in RAID1 (the total disk capacity after RAID setup must be greater than or equal to 50 GB), installation path: /var/lib/etcd ¡ The RAID controller must have 1 GB cache memory, support power fail protection, and have a supercapacitor in place. NICs: 1 × 10Gbps |
Carrier device simulation analyzer deployment
Simulation analyzer hardware requirements (Analyzer E6502 or later)
This section describes the server requirements for the analyzer to use the simulation feature. To deploy the simulation analyzer, add two additional network adapter ports to manage DTN physical host and collect simulated VSR device data, respectively. The device simulation feature requires an additional server to act as the DTN physical host. The DTN physical host is used to run the simulated VSR devices, and the required resources are shown in Table 103.
Table 77 DTN physical host deployment (x86-64(Intel64/AMD64))
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
DTN physical host |
A server can manage a maximum of 30 simulation devices. Number of servers required = number of simulation devices/30 |
x86-64(Intel64/AMD64) CPUs supporting VX-T/VX-D, 16 cores, 2.0 GHz or above |
128 GB or above |
System disk: 600 GB or above |
Support 1 to 10Gbps bandwidth A minimum of three network adapter ports |
|
DTN physical host |
A server can manage a maximum of 60 simulation devices. Number of servers required = number of simulation devices/60 |
x86-64(Intel64/AMD64) CPUs supporting VX-T/VX-D, 20 cores, 2.2 GHz or above |
256 GB or above |
|
|
Simulation analyzer hardware requirements (Analyzer versions earlier than E6502)
This section describes the server requirements for the analyzer to use the simulation feature. To deploy the simulation analyzer, add an additional network adapter port to manage DTN physical host. The device simulation feature requires additional servers to act as the DTN physical host and simulation analyzer. The DTN physical host is used to run the simulated VSR devices, and the required resources are shown in Table 104. The simulation analyzer server is used to manage the VSR devices on the DTN physical host, and the required resources are shown in Table 105.
Table 78 DTN physical host deployment (x86-64(Intel64/AMD64))
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
DTN physical host |
A server can manage a maximum of 30 simulation devices. Number of servers required = number of simulation devices/30 |
x86-64(Intel64/AMD64) CPUs supporting VX-T/VX-D, 16 cores, 2.0 GHz or above |
128 GB or above |
System disk: 600 GB or above |
Support 1 to 10Gbps bandwidth A minimum of three network adapter ports |
|
DTN physical host |
A server can manage a maximum of 60 simulation devices. Number of servers required = number of simulation devices/60 |
x86-64(Intel64/AMD64) CPUs supporting VX-T/VX-D, 20 cores, 2.2 GHz or above |
256 GB or above |
Table 79 Simulation analyzer physical host deployment (x86-64(Intel64/AMD64))
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Simulation analyzer |
200 |
16 cores (total physical cores), 2.0GHz or above |
160G or higher |
System disk: 1.92 TB or above after RAID setup Data disk: 1 TB or above after RAID setup ETCD disk: 50 GB or above after RAID setup, installation path: /var/lib/etcd |
Network adapter ports: · Non-bonding mode: 2 × 10 Gbps or above ports. · Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
Optional component deployment (x86-64(Intel64/AMD64))
Table 80 Hardware requirements for optional component deployment in single-node mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
Lite mode: Network devices: 200 Network performance collection instances: 10000 Online users: 5 Concurrent users: 1 Trap rate ≤ 5/sec |
3 |
12 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the component. To configure network management services based on network elements, add additional resources. No additional disk resources are required as they have been reserved for the scenario. |
|
General mode (default): Network devices: 2000 Network performance collection instances: 50000 Online users: 50 Concurrent users: 10 Trap rate ≤ 40/sec |
4 |
36 |
|||
|
Syslog |
0.75 |
8 |
Provides the syslog function. |
Optional. Require additional hardware resources. |
|
|
Wvas (QoSM) |
1.5 |
4.5 |
QoS template, route policy template |
Optional. Require additional hardware resources. |
|
|
Slice (FlexE) |
1 |
2.5 |
Network slicing |
Optional. Require additional hardware resources. |
|
|
wBGP |
1 |
2 |
BGP Flowspec & BGP SRv6 policy southbound protocol |
Optional. Require additional hardware resources. |
|
Table 81 Hardware requirements for optional component deployment in cluster mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
General mode (default): Network devices: 200 Network performance collection instances: 10000 Online users: 5 Concurrent users: 1 Trap rate ≤ 40/sec |
2 |
24 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the component. To configure network management services based on network elements, add additional resources. No additional disk resources are required as they have been reserved for the scenario. |
|
General mode (default): Network devices: 2000 Network performance collection instances: 200000 Online users: 100 Concurrent users: 20 Trap rate ≤ 80/sec |
3 |
56 |
|||
|
Syslog |
0.75 |
8 |
Provides the syslog function. |
Optional. Require additional hardware resources. |
|
|
Wvas (QoSM) |
1.5 |
4.5 |
QoS template, route policy template |
Optional. Require additional hardware resources. |
|
|
Slice (FlexE) |
1 |
2.5 |
Network slicing |
Optional. Require additional hardware resources. |
|
|
wBGP |
1 |
2 |
BGP Flowspec & BGP SRv6 policy southbound protocol |
Optional. Require additional hardware resources. |
|
Optional component deployment (x86-64 Hygon server/ARM Kunpeng server/ARM Phytium server)
Table 82 Hardware requirements for optional component deployment in single-node mode (x86-64 Hygon server/ARM Kunpeng server/ARM Phytium server)
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
Lite mode: Network devices: 200 Network performance collection instances: 10000 Online users: 5 Concurrent users: 1 Trap rate ≤ 5/sec |
6 |
12 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the component. To configure network management services based on network elements, add additional resources. No additional disk resources are required as they have been reserved for the scenario. |
|
General mode (default): Network devices: 2000 Network performance collection instances: 50000 Online users: 50 Concurrent users: 10 Trap rate ≤ 40/sec |
8 |
36 |
|||
|
Syslog |
1.5 |
8 |
Provides the syslog function. |
Optional. Require additional hardware resources. |
|
|
Wvas (QoSM) |
3 |
4.5 |
QoS template, route policy template |
Optional. Require additional hardware resources. |
|
|
Slice (FlexE) |
2 |
2.5 |
Network slicing |
Optional. Require additional hardware resources. |
|
|
wBGP |
2 |
2 |
BGP Flowspec & BGP SRv6 policy southbound protocol |
Optional. Require additional hardware resources. |
|
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
General mode (default): Network devices: 200 Network performance collection instances: 10000 Online users: 5 Concurrent users: 1 Trap rate ≤ 40/sec |
4 |
24 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the component. To configure network management services based on network elements, add additional resources. No additional disk resources are required as they have been reserved for the scenario. |
|
General mode (default): Network devices: 2000 Network performance collection instances: 200000 Online users: 100 Concurrent users: 20 Trap rate ≤ 80/sec |
6 |
56 |
|||
|
Syslog |
1.5 |
8 |
Provides the syslog function. |
Optional. Require additional hardware resources. |
|
|
Wvas (QoSM) |
3 |
4.5 |
QoS template, route policy template |
Optional. Require additional hardware resources. |
|
|
Slice (FlexE) |
2 |
2.5 |
Network slicing |
Optional. Require additional hardware resources. |
|
|
wBGP |
2 |
2 |
BGP Flowspec & BGP SRv6 policy southbound protocol |
Optional. Require additional hardware resources. |
|
Hardware requirements for AD-WAN branch scenarios
Hardware requirements for AD-WAN branch access network deployment (controller E6804 and later)
Both single-node deployment and cluster deployment are supported. The controller can be deployed on both physical servers and VMs.
For standalone controller deployment, you must manually configure the parameters. For converged deployment, configure the parameters based on the component requirements. For more information, see AD-WAN Branch Unified Platform and Components Deployment Guide.
Configure the parameters in standalone deployment mode as follows:
· Single-node/cluster deployment within 200: Specify the Unified Platform Lite mode, set the kafka limit value to 2G on Matrix, and set the websocket limit to 3G.
· Single-node deployment with 200 to 2000: Set the kafka limit value to 6G on Matrix, and set the websocket limit to 11G.
· Cluster deployment with 200 to 2000: Set the kafka limit value to 6G on Matrix, and set the websocket limit to 6G.
|
|
NOTE: · For Unified Platform versions earlier than E0706 (including E06xx), the etcd partition requires a separate physical drive. · For Unified Platform E0706 and later versions, the etcd partition can share a physical drive with other partitions. As a best practice, use a separate physical drive for the etcd partition. · The configuration in this section does not involve that for the network management services. To use complete network management services, add additional resources for them. |
Deployment on physical servers
General disk requirements
This section introduces the general disk configuration of physical servers, including recommended RAID mode, disk type, and performance requirements. For specific disk capacity requirements, see the configuration requirements in the corresponding scenarios.
Table 84 General requirements
|
Item |
Requirements |
||
|
General disk requirements: Configure RAID 1, RAID 5, or RAID 10. Use SSDs or 7.2K RPM or above HDDs. IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. For HDDs, the RAID controller must have 1 GB cache memory, support power fail protection, and have a supercapacitor in place. |
System disk |
SSD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2 × 1920 GB disks · RAID 5: 3 × 960 GB or 5 × 480 GB disks · RAID 10: 4 × 960 GB or 8 × 480 GB disks |
|
HDD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2 × 1920 GB disks · RAID 5: 3 × 1200 GB or 5 × 600 GB disks · RAID 10: 4 × 1200 GB or 8 × 600 GB disks |
||
|
ETCD disk |
SSD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2 × 480GB disks |
|
|
HDD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2 × 600GB disks |
||
|
Data disk |
A minimum of 8 TB disk capacity after RAID setup. Recommended configuration: · RAID 5: A minimum of three disks of the same model A minimum of 12 TB disk capacity after RAID setup. Recommended configuration: · RAID 5: A minimum of five disks of the same model A minimum of 24 TB disk capacity after RAID setup. Recommended configuration: · RAID 5: A minimum of seven disks of the same model |
||
|
|
NOTE: For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software.
|
Standalone deployment of the branch controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 85 Hardware requirements for branch controller deployment in cluster mode (including Unified Platform)
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller |
200 (Lite mode) |
Intel Xeon V3 or higher, 12 or more cores 2.0 GHz or above |
96 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
· Non-bonding mode: 1 × 1Gbps network port. · Bonding mode (mode 2 or mode 4 is recommended): 2 × 1 Gbps or above network ports, forming a Linux bonding interface. |
|
Controller |
2000 |
Intel Xeon V3 or higher, 16 or more cores 2.0 GHz or above |
128 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
Table 86 Hardware requirements for branch controller deployment in single-node mode (including Unified Platform)
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller |
200 (Lite mode) |
Intel Xeon V3 or higher, 12 or more cores 2.0 GHz or above |
96 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
· Non-bonding mode: 1 × 1Gbps network port. · Bonding mode (mode 2 or mode 4 is recommended): 2 × 1 Gbps or above network ports, forming a Linux bonding interface. |
|
Controller |
2000 |
Intel Xeon V3 or higher, 16 or more cores 2.0 GHz or above |
128 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
Converged deployment of the branch controller
This section describes the hardware requirements for converged deployment of SeerAnalyzer E62xx or later and the controller.
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
The following converged deployment modes are available:
· Converged deployment of the branch controller and analyzer in single-node mode. For the hardware requirements, see Table 113. To deploy the security controller, add additional hardware resources as required in "Optional component deployment."
· Converged deployment of the branch controller and analyzer in three-node cluster mode. In this mode, deploy the controller and analyzer on three master nodes. For hardware requirements, see Table 114. To deploy the security controller, add additional hardware resources for the 3 master nodes as required in "Optional component deployment."
· Converged deployment of the branch controller and analyzer in 3+N cluster mode. In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. Table 115 shows the hardware requirements for the branch controller in this mode. Table 116 shows the hardware requirements for the analyzer in 3+1 cluster mode. Table 117 shows the hardware requirements for the analyzer in 3+3 cluster mode. To deploy the security controller, add additional hardware resources for the 3 master nodes as required in "Optional component deployment." The security controller must be deployed on master nodes.
|
|
NOTE: In 3+N cluster deployment mode, the analyzer is deployed on worker nodes. In this scenario, the hardware requirements vary by analyzer version. This document describes only the hardware requirements for SeerAnalyzer E62xx and later. For hardware requirements of SeerAnalyzer history versions, see "Hardware requirements for SeerAnalyzer history versions." |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller + analyzer |
200 |
Intel Xeon V3 or higher, 28 or more cores 2.0 GHz or above |
192 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
|
|
IMPORTANT: In single-node deployment mode, follow these restrictions and guidelines: · The single-node deployment mode cannot provide redundancy, and device failures will cause service interruption. As a best practice, configure a remote backup server. · If the controller and analyzer are deployed in a converged manner on a single node, you cannot scale up the system to a cluster. |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller + analyzer |
200 |
Intel Xeon V3 or higher, 28 or more cores 2.0 GHz or above |
160 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
|
Controller + analyzer |
2000 |
Intel Xeon V3 or higher, 36 or more cores 2.0 GHz or above |
256 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. |
|
Components |
Device quantity |
CPU |
Memory |
Drive |
Network ports |
|
Controller |
200 |
Intel Xeon V3 or higher, 20 or more cores 2.0 GHz or above |
128 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
· Non-bonding mode: 2 × 10Gbps network ports. · Bonding mode (mode 2 or mode 4 is recommended): 4 × 10 Gbps or above network ports, each two forming a Linux bonding interface. |
|
Controller |
2000 |
Intel Xeon V3 or higher, 24 or more cores 2.0 GHz or above |
160 GB or more |
System drive: 1.92 TB or above ETCD drive: 50 GB or above |
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
1 |
· CPU: 24 cores (total physical cores), 2.0 GHz · Memory: 160 GB. · System drive: 1.92 TB after RAID setup · Data drive: 2 TB after RAID setup. 2 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd. · Network ports: ¡ Non-bonding mode: 2 × 10Gbps Ethernet ports. ¡ Bonding mode (recommended mode: mode 2 or mode 4): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
200 |
|
Node configuration |
Maximum resources that can be managed |
||
|
Node name |
Node quantity |
Minimum single-node requirements |
|
|
Analyzer node |
3 |
· CPU: 18 cores (total physical cores), 2.0 GHz · Memory: 140 GB. · System drive: 1.92 TB after RAID setup · Data drive: 4 TB after RAID setup. 3 or more drives of the same model are required. · ETCD drive: 50 GB (after RAID setup). Installation path: /var/lib/etcd · Network ports: ¡ Non-bonding mode: 2 × 10Gbps Ethernet ports. ¡ Bonding mode (recommended mode: mode 2 or mode 4): 4 × 10 Gbps network ports, each two forming a Linux bonding interface. |
2000 |
Hardware requirements for deployment on VMs
SD-WAN supports deployment on VMs.
For deployment on a VM, if the host where the VM resides is enabled with hyperthreading, the number of required vCPUs is twice the number of CPU cores required for deployment on a physical server. If the host is not enabled with hyperthreading, the number of required vCPUs is the same as the number of CPU cores required for deployment on a physical server. The memory and drive resources required for deployment on a VM are the same as those required for deployment on a physical server. As a best practice, enable hyperthreading.
For deployment on a VM, follow these restrictions and guidelines:
· VM deployment requires VMware ESXi 6.7.0, H3C_CAS-E0706, or their later versions.
· 4k block file read/write IOPS ≥ 10000. When shared storage is used, the IOPS performance degrades as the number of users increases. Make sure the IOPS performance meets the requirements.
· VM migration is not supported.
· Do not deploy the VMs in the cluster on the same host.
· Do not enable overcommitment on the VMs. To ensure system stability, reserve resources for the controller VM.
· Do not set the simplified mode for the VM storage volumes.
· The VM cluster nodes must be installed separately. You cannot clone a node and then modify its IP.
· As a best practice, install Unified Platform on a VM through mounting an ISO image rather than building a hard disk image.
· Do not deploy virtual cluster nodes and other IO-intensive VMs on the same server.
· When you specify an etcd partition, make sure the data disks do not use the same storage pool. If you have an SSD storage pool, as a best practice, use the SSD storage pool to allocate an etcd partition.
· To deploy the controller on a VMware VM, you must enable promiscuous mode and pseudo-transmission on the host where the VM resides.
|
|
IMPORTANT: · As a best practice, use physical servers for deployment if the network contains more than 200 devices. · If the compatible Unified Platform version is earlier than E0706 (including E06xx), the ETCD partition must use a separate physical disk. If the compatible Unified Platform version is E0706 or later, the ETCD partition can share a physical disk with other partitions. As a best practice, use a separate physical disk for the ETCD partition. |
Optional component deployment
Table 92 Hardware requirements for optional component deployment in single-node mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
Lite mode: Network devices: 200 Network performance collection instances: 10000 Online users: 5 Concurrent users: 1 Trap rate ≤ 5/sec |
3 |
12 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the RES component. To use complete network management services, you must add additional resources. |
|
General mode (default): Network devices: 2000 Network performance collection instances: 50000 Online users: 50 Concurrent users: 10 Trap rate ≤ 40/sec |
4 |
36 |
|||
|
Syslog |
0.75 |
8 |
Syslog |
Optional. Require additional hardware resources. |
|
|
SSM |
1 |
3 |
APR signature library management |
Optional. Require additional hardware resources. |
|
|
MBM |
2 |
10 |
Mobile branch mgmt |
Optional. Require additional hardware resources. |
|
|
SeerEngine_SECURITY |
2 |
32 |
Security service and signature library |
Optional. Require additional hardware resources. A maximum of 200 devices and 60000 policies are supported. |
|
Table 93 Hardware requirements for optional component deployment in cluster mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
||
|
Network |
General mode (default): · Network devices: 200 · Network performance collection instances: 10000 · Online users: 5 · Concurrent users: 1 Trap rate ≤ 40/sec |
2 |
24 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the RES component. |
|
|
General mode (default): Network devices: 2000 Network performance collection instances: 200000 Online users: 100 Concurrent users: 20 Trap rate ≤ 80/sec |
3 |
56 |
||||
|
Syslog |
0.75 |
8 |
Syslog |
Optional. Require additional hardware resources. |
||
|
SSM |
1 |
3 |
APR signature library management |
Optional. Require additional hardware resources. |
||
|
MBM |
2 |
10 |
Mobile branch mgmt |
Optional. Require additional hardware resources. |
||
|
SeerEngine_SECURITY |
A maximum of 200 devices and 60000 policies are supported. |
4 |
32 |
Security service and signature library |
Optional. Require additional hardware resources. In a 3-node cluster, the component can be deployed only on the master node. |
|
|
A maximum of 2000 devices and 240000 policies are supported. |
8 |
|||||
Hardware requirements for AD-WAN branch access network deployment (controller E6803 or earlier)
Both single-node deployment and cluster deployment are supported. The controller can be deployed on both physical servers and VMs.
For standalone controller deployment, you must manually configure the parameters. For converged deployment, configure the parameters based on the component requirements. For more information, see the deployment guide.
Configure the parameters in standalone deployment mode as follows:
· Single-node/cluster deployment within 200: Set the kafka limit value to 2G on Matrix, and set the websocket limit to 3G.
· Single-node deployment with 200 to 2000: Set the kafka limit value to 6G on Matrix, and set the websocket limit to 11G.
· Cluster deployment with 200 to 2000: Set the kafka limit value to 6G on Matrix, and set the websocket limit to 6G.
|
|
NOTE: If the compatible Unified Platform version is earlier than E0706 (including E06xx), the ETCD partition must use a separate physical disk. If the compatible Unified Platform version is E0706 or later, the ETCD partition can share a physical disk with other partitions. As a best practice, use a separate physical disk for the ETCD partition. The configuration in this section does not involve that for the network management services. To use complete network management services, add additional resources for them. |
Server requirements
General disk requirements
This section introduces the general disk configuration of physical servers, including recommended RAID mode, disk type, and performance requirements. For specific disk capacity requirements, see the configuration requirements in the corresponding scenarios.
Table 94 General requirements
|
Item |
Requirements |
||
|
General disk requirements: Configure RAID 1, RAID 5, or RAID 10. Use SSDs or 7.2K RPM or above HDDs. IOPS ≥ 10000. See the check result of the Unified Platform env_check.sh script. For HDDs, the RAID controller must have 1 GB cache memory, support power fail protection, and have a supercapacitor in place. |
System disk |
SSD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2*1920GB · RAID5: 3 × 960 GB or 5 × 480 GB disks · RAID10: 4 × 960 GB or 8 × 480 GB disks |
|
HDD |
A minimum of 1920 GB disk capacity after RAID setup. Recommended configuration: · RAID1: 2*1920GB · RAID5: 3 × 1200 GB or 5 × 600 GB disks · RAID10: 4 × 1200 GB or 8 × 600 GB disks |
||
|
ETCD disk |
SSD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2*480GB |
|
|
HDD |
A minimum of 50 GB disk capacity after RAID setup. Recommended configuration: RAID1: 2*600GB |
||
|
Data disk |
A minimum of 8 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of three disks of the same model A minimum of 12 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of five disks of the same model A minimum of 24 TB disk capacity after RAID setup. Recommended configuration: · RAID5: A minimum of seven disks of the same model |
||
|
|
NOTE: For Unified Platform E0714 and later versions, you can set up RAID through software after deploying operating system. Make sure the disk capacity after RAID setup meets the IOPS performance requirements. For Unified Platform E0713 and earlier versions, you cannot set up RAID through software. |
Standalone deployment of branch controller
This section describes only the hardware configuration for the required Unified Platform application packages and the controller software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guide of the controller.
Table 95 Branch controller deployment in cluster mode (including Unified Platform)
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller |
200 |
Intel Xeon V3 series or later, total cores ≥ 12 Clock speed: 2.0 GHz or above |
128 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
Non-bonding mode: 1 × 1 Gbps or above port. Bonding mode: 2 × 1 Gbps or above ports, forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller |
2000 |
Intel Xeon V3 series or later, total cores ≥ 16 Clock speed: 2.0 GHz or above |
128 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
Table 96 Branch controller deployment in single-node mode (including Unified Platform)
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller |
200 |
Intel Xeon V3 series or later, total cores ≥ 12 Clock speed: 2.0 GHz or above |
128 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
Non-bonding mode: 1 × 1 Gbps or above port. Bonding mode: 2 × 1 Gbps or above ports, forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller |
2000 |
Intel Xeon V3 series or later, total cores ≥ 16 Clock speed: 2.0 GHz or above |
128 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
Converged deployment of the branch controller
This section describes the hardware requirements for converged deployment of the controller and analyzer E62xx or above.
This section describes only the hardware configuration for the required Unified Platform application packages, controller software, and analyzer software. To deploy optional application packages, add hardware configuration for them. To identify whether an application package is required or optional, see the installation guides of the controller and analyzer.
The following modes are available for converged deployment of branch controller, security controller, and analyzer:
· Converged deployment of branch controller and analyzer in single-node mode: The hardware requirements are shown in Table 123. To deploy also the security controller, add additional hardware resources as shown in "Optional component deployment."
· Converged deployment of the branch controller and analyzer in 3-node cluster mode: In this mode, deploy the branch controller and analyzer in a converged manner on the three master nodes in a Unified Platform cluster. The hardware requirements are shown in Table 124. To deploy also the security controller, add additional hardware resources for the three master nodes as shown in"Optional component deployment."
· Converged deployment of the branch controller and analyzer in 3+N cluster mode: In this mode, deploy the controller on three master nodes and deploy the analyzer on N (1 or 3) worker nodes. Table 125 shows the hardware requirements for the branch controller in this mode. Table 126 shows the hardware requirements for the analyzer in 3+1 cluster mode. Table 127 shows the hardware requirements for the analyzer in 3+3 cluster mode. To deploy the security controller, add additional hardware resources for the 3 master nodes as required in "Optional component deployment." The security controller must be deployed on master nodes.
|
|
NOTE: In 3+N cluster deployment mode, the analyzer is deployed on the worker node. The hardware requirements for analyzer vary by software version. This section describes the hardware requirements for analyzer E62xx or above. For hardware requirements of history versions, see "Hardware requirements for SeerAnalyzer history versions." |
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
controller+analyzer |
200 |
Intel Xeon V3 series or later, total cores ≥ 28 Clock speed: 2.0 GHz or above |
192 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above Data disk: 2 TB or above after RAID setup by using two or more disks of the same model |
Non-bonding mode: 2 × 10 Gbps or above ports. Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
|
IMPORTANT: In single-node deployment mode, follow these restrictions and guidelines: · No redundancy node exists in single-node mode, and a failure will lead to service interruption. As a best practice, configure a remote backup server. · For converged deployment of the controller and analyzer in single-node mode, you cannot change the deployment mode to cluster mode. |
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller + analyzer |
200 |
Intel Xeon V3 series or later, total cores ≥ 28 Clock speed: 2.0 GHz or above |
160 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above Data disk: 2 TB or above after RAID setup by using two or more disks of the same model |
Non-bonding mode: 2 × 10 Gbps and above bandwidth. Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller + analyzer |
2000 |
Intel Xeon V3 series or later, total cores ≥ 36 Clock speed: 2.0 GHz or above |
256 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above Data disk: 4 TB or above after RAID setup by using three or more disks of the same model |
|
Application |
Device quantity |
CPU |
Memory |
Disk |
NIC |
|
Controller |
200 |
Intel Xeon V3 series or later, total cores ≥ 20 Clock speed: 2.0 GHz or above |
158 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
Non-bonding mode: 2 × 10 Gbps or above ports. Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
|
Controller |
2000 |
Intel Xeon V3 series or later, total cores ≥ 24 Clock speed: 2.0 GHz or above |
158 GB or above |
System disk: 1.92 TB or above ETCD disk: 50 GB or above |
|
Hardware configuration |
Scale |
||
|
Node name |
Node quantity |
Single-node requirements |
|
|
Analyzer node |
1 |
CPU: 24 cores (total physical cores), 2.0 GHz or above Memory: 160 GB System disk: 1.92 TB after RAID setup Data disk: 2 TB or above after RAID setup by using two or more disks of the same model ETCD disk: 50 GB or above after RAID setup, installation path: /var/lib/etcd Network adapter ports: Non-bonding mode: 2 × 10 Gbps and above bandwidth. Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
200 |
|
Hardware configuration |
|
|||
|
Node name |
Node quantity |
Single-node requirements |
Scale |
|
|
Analyzer node |
3 |
CPU: 18 cores (total physical cores), 2.0 GHz or above Memory: 140 GB System disk: 1.92 TB after RAID setup Data disk: 4 TB or above after RAID setup by using three or more disks of the same model ETCD disk: 50 GB or above after RAID setup, installation path: /var/lib/etcd Network adapter ports: Non-bonding mode: 2 × 10 Gbps and above bandwidth. Bonding mode: 4 × 10 Gbps or above ports, each two forming a Linux bond interface. Bond mode 2 or 4 is recommended. |
2000 |
|
Deployment on VMs
For VM-based deployment, if the virtualization platform server has the hyper-threading mode enabled, the number of vCPUs must be twice the number of CPU cores required by physical server-based deployment. If the hyper-threading mode is not enabled, the number of vCPUs must be the same as that required by physical server-based deployment, and the memory and disk configuration can follow the requirements by physical server-based deployment. As a best practice, enable hyper-threading for the servers.
For deployment on a VM, follow these restrictions and guidelines:
· VM deployment requires VMware ESXi 6.7.0, H3C_CAS-E0706, or their later versions.
· 4k block file read/write IOPS ≥ 10000. When shared storage is used, the IOPS performance degrades as the number of users increases. Make sure the IOPS performance meets the requirements.
· VM migration is not supported.
· Do not deploy the VMs in the cluster on the same host.
· Do not enable overcommitment on the VMs. To ensure system stability, reserve resources for the controller VM.
· Do not set the simplified mode for the VM storage volumes.
· The VM cluster nodes must be installed separately. You cannot clone a node and then modify its IP.
· As a best practice, install Unified Platform on a VM through mounting an ISO image rather than building a hard disk image.
· Do not deploy virtual cluster nodes and other IO-intensive VMs on the same server.
· When you specify an etcd partition, make sure the data disks do not use the same storage pool. If you have an SSD storage pool, as a best practice, use the SSD storage pool to allocate an etcd partition.
· To deploy the controller on a VMware VM, you must enable promiscuous mode and pseudo-transmission on the host where the VM resides.
|
|
IMPORTANT: · As a best practice, deploy the controller on physical servers if the network contains more than 200 devices. · If the compatible Unified Platform version is earlier than E0706 (including E06xx), the ETCD partition must use a separate physical disk. If the compatible Unified Platform version is E0706 or later, the ETCD partition can share a physical disk with other partitions. As a best practice, use a separate physical disk for the ETCD partition. |
Optional component deployment
Table 102 Hardware requirements for optional component deployment in single-node mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
|
|
Network |
Lite mode: · Network devices: 200 · Network performance collection instances: 10000 · Online users: 5 · Concurrent users: 1 Trap rate ≤ 5/sec |
3 |
12 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the RES component. To use complete network management services, you must add additional resources. |
|
General mode (default): · Network devices: 2000 · Network performance collection instances: 50000 · Online users: 50 · Concurrent users: 10 Trap rate ≤ 40/sec |
4 |
36 |
|||
|
Syslog |
0.75 |
8 |
Syslog |
Optional. Require additional hardware resources. |
|
|
SSM |
2 |
4 |
APR signature library management |
Optional. Require additional hardware resources. |
|
|
MBM |
2 |
10 |
Mobile branch mgmt |
Optional. Require additional hardware resources. |
|
|
SeerEngine_SECURITY |
4 |
32 |
Security service and signature library |
Optional. Require additional hardware resources. A maximum of 200 devices and 60000 policies are supported. |
|
Table 103 Hardware requirements for optional component deployment in cluster mode (x86-64(Intel64/AMD64))
|
Application installation package |
CPU (physical cores) |
Memory (GB) |
Description |
Remarks |
||
|
Network |
General mode (default): · Network devices: 200 · Network performance collection instances: 10000 · Online users: 5 · Concurrent users: 1 Trap rate ≤ 40/sec |
2 |
24 |
Network resource, network performance, network topology, and iCC |
No additional resources are required if you install only the RES component. |
|
|
General mode (default): Network devices: 2000 Network performance collection instances: 200000 Online users: 100 Concurrent users: 20 Trap rate ≤ 80/sec |
3 |
56 |
||||
|
Syslog |
0.75 |
8 |
Syslog |
Optional. Require additional hardware resources. |
||
|
SSM |
1 |
3 |
APR signature library management |
Optional. Require additional hardware resources. |
||
|
MBM |
2 |
10 |
Mobile branch mgmt |
Optional. Require additional hardware resources. |
||
|
SeerEngine_SECURITY |
A maximum of 200 devices and 60000 policies are supported. |
4 |
32 |
Security service and signature library |
Optional. Require additional hardware resources. In a 3-node cluster, the component can be deployed only on the master node. |
|
|
A maximum of 2000 devices and 240000 policies are supported. |
8 |
|||||
Hardware requirements for license server deployment
Table 148 shows the hardware requirements for license server deployment.
Table 104 Hardware requirements
|
CPU architecture |
CPU kernel |
Memory |
Drive size |
NIC |
Remarks |
|
x86-64(Intel64/AMD64) |
16 cores, 2.0GHz or above |
32 GB or above |
64GB or above (system partition where the root directory resides) |
Support 1 to 10Gbps bandwidth |
Recommended configuration This configuration supports the maximum allowed number of clients. |
|
x86-64(Intel64/AMD64) |
4 cores or above 2.0GHz or above |
4 GB or above |
64GB or above (system partition where the root directory resides) |
Support 1 to 10Gbps bandwidth |
Minimum configuration This configuration supports up to 20 clients. |
|
|
NOTE: · To ensure stable operation of the license server, deploy the license server on a physical server instead of a VM. · When the license server is deployed with the controller in a converged manner, you do not need to reserve hardware resources for the license server. |