Explore Cisco
How to Buy

Have an account?

  •   Personalized content
  •   Your products and support

Need an account?

Create an account

Planning Guide and System Requirements for Cisco WebEx Meetings Server Release 2.8

Book Contents
Find Matches in This Book
Available Languages
Download Options

Book Title

Planning Guide and System Requirements for Cisco WebEx Meetings Server Release 2.8

Chapter Title

Webex System Requirements

Results

Updated:
May 26, 2020

Chapter: Webex System Requirements

Webex System Requirements

General System Requirements

Cisco Webex Meetings Server (CWMS) is compatible with Cisco UCS servers that meet or exceed the specifications presented in this section.


Important

When you perform an upgrade to a major release of CWMS, such as to Release 2.0 or Release 2.5 from Release 1.x, the ESXi hosts (Cisco UCS server) where the Admin virtual machine is located require a minimum of 1.5-TB of free disk space. Refer to the section in this document that describes the different size user systems that begin with the 50-user System. During an upgrade, there are two sets of virtual machines on your network at the same time; the original virtual machines running Release 1.x and the upgrade virtual machines to support the new release. For more details, see the "Upgrading the System" section in the CWMS Administration Guide at http://www.cisco.com/en/US/products/ps12732/prod_installation_guides_list.html.

Module

Requirements Notes

Host server and processors

  • Cisco UCS C-series rack server or equivalent B-series blade server.

  • AES-NI instruction set support.

  • 2.4 GHz or faster processor clock speed.

Note 

Third-party hardware is not supported.

Network interfaces

The NICs between the ESXi hosts (for the Cisco Webex Meetings Server virtual machines) and the Ethernet switch (not to the external network interface).

  • Minimum 1 physical NIC for a non-redundant configuration. See the 50-user System section for special requirements where the Internet Reverse Proxy (IRP) and Admin virtual machine are sharing a host.

  • Redundant configurations must have all NIC interfaces duplicated (teamed or bonded) and connected to an independent switching fabric.

  • An additional NIC for the VMware management network (optional).

1Internal (DAS) Storage for ESXi hosts where internal virtual machines are deployed

  • Minimum of 4 drives in a RAID-10 or RAID-5 configuration

  • Minimum of 1.5 TB usable storage for new system deployments or upgrades.

  • When you upgrade to CWMS Release 2.X from Release 1.X, the ESXi hosts each require from 172 to 1118-GB free disk space depending on the size of your system and the virtual machines. For more information, see Minimum Hardware Requirements.

  • Optional second array for ESXi

Note 

The virtual machines must use thick provisioning for storage.

Internal (DAS) storage for ESXi hosts where IRP virtual machines are deployed

  • Minimum of 2 drives in a RAID-1 configuration

  • Minimum of 300-GB usable storage

  • Can use the same configurations as for the internal virtual machines

Note 

The virtual machines must use thick provisioning for storage.

SAN storage

  • Can be used as a substitute for DAS. (We recommend allocating of the same amount of storage space.)

  • B-series blade servers have only two hard disk drives. If you are using Cisco UCS B-series blade servers, upgrading requires the use of SAN storage. SAN storage meets the requirement to have 4 hard disk drives in either a RAID 5 or RAID 10 configuration.

  • Recommended only for deployments where the support staff has experience monitoring and tuning SAN performance.

    Note 

    You take responsibility for adding storage for new VMware requirements and future growth of the system.

  • Fiber Channel (FC) or Fiber Channel over 10-GB Ethernet (FCoE) only.

  • Performance requirements are the same as for DAS.

Hypervisor

ESXi versions and vSphere licenses are described in the Minimum Hardware Requirements section.

VMware Vsphere is required and the only product supported; other hypervisor products are not supported.

One VMware license per processor socket.

For more information about vSphere licenses, see Minimum Hardware Requirements.

VMware vCenter Server or vCenter Server appliance versions: 5.0, 5.0 Update 1, 5.0 Update 2, 5.1, 5.1 Update 1, 5.5. (For security reasons, 5.1 must be version 1900470 or later.)

Coresidency:

  • vCenter can be coresident with CWMS, providing the processor and memory requirements are added to the system requirements.

  • vCenter coresident configurations are supported only for 50-user and 250-user systems.

  • Coresidency with Cisco Unified Communications products on the same physical ESXi host is not supported.

  • Coresidency with non-CWMS virtual machines on the same physical ESXi host is not supported.

Advanced VMware vSphere features such as Distributed Resource Scheduler (DRS), Cloning, Fault Tolerance (FT), and vMotion or Storage vMotion are not supported.

Email server

  • Fully qualified domain name (FQDN) of the mail server that the system uses to send emails.

  • Port number—default value of the SMTP port number is 25 or 465 (secure SMTP port number).

  • To use a TLS-enabled email server with third-party certificates, you must import the certificates into your system. For more information, See "Managing Certificates" in the administration guide at http://www.cisco.com/en/US/products/ps12732/prod_installation_guides_list.html.

Licenses

MDC licenses—An MDC license is required for each data center in a multi-data center system. A MDC license is not required for a system with a single data center.

Host license—Each user that shall host a meeting must have a Host license to start a Webex Web, Webex Audio, Blast Dial meeting, or Personal Conferencing. For more information about Host licenses, see About Host Licenses.

1 If your organization has expertise in managing a storage area network (SAN), we recommend SAN over direct attached storage (DAS). SANs can be more reliable than local disk arrays.

Cisco WebEx Meetings Server Best Practices

The following is a list of best practices that you should refer to when configuring and maintaining your Cisco WebEx Meetings Server system:

  • Power your virtual machine hosts by using an uninterruptible power source (UPS) to minimize power interruptions. Repeated power failures can damage host systems and virtual machines.

  • Always put your system into maintenance mode before shutting down a guest operating system.

  • For scheduled events and other situations that require a system shutdown, gracefully shut down your virtual machines by shutting down the guest operating system.

  • The system is designed to repair itself when necessary and rebooting can interrupt this process. We do not recommend that you reboot your system to fix it. If your system is in an unhealthy state, contact the Cisco TAC. Power off your system only when instructed to do so or during scheduled events such as data center maintenance.

  • Configure network redundancy to minimize network failures. Refer to "Adding a High Availability System" in the Cisco WebEx Meetings Server Administration Guide for more information.

  • Using snapshots on your virtual machines can impair system performance in ways that affect user experience even when the system is otherwise lightly loaded.

  • If your system is having problems, make sure that you check your VMware VCenter environment to determine if conditions in VCenter or the network are causing the problem.

  • Configure high availability to increase the probability that your system can continue to operate if a failure occurs.

  • If you have a high-availability system and your secondary system fails, you can repair it by removing the existing secondary system (refer to "Removing a High Availability System" in your Cisco WebEx Meetings Server Administration Guide) and adding a new secondary system (refer to "Adding a High Availability System" in your Cisco WebEx Meetings Server Administration Guide). If the primary system on a high-availability system fails, you cannot repair it using this procedure. We recommend that you restore your primary system using the disaster recovery procedure and then add a new secondary system. Until you add a new secondary system your deployment is operating without full redundancy. This procedure helps prevent unplanned outages if any of your secondary virtual machines fails. Refer to "Using the Disaster Recovery Feature" in the Cisco WebEx Meetings Server Administration Guide for more information.

  • Since your system only keeps the latest system backup on the NFS and removes previous instances every day, we recommend that you keep several recent backups on other media.


    Restriction

    Do not manually create files or directories in the NFS share used by Cisco WebEx Meetings Server, as it runs various scripts on NFS files and directories. The NFS storage server must be for the exclusive use of Cisco WebEx Meetings Server.

  • Use your dashboard to monitor the health status of the NFS, CPU, and storage. We recommend enabling dashboard alarms for storage and CPU.

  • If you plan to use directory integration, refer to the Configuring Directory Integration section in the "Managing Users" chapter of the Cisco WebEx Meetings Server Administration Guide for more information.

  • When using Cisco WebEx Meetings Server, the related SIP trunk on CUCM in the CallManager service interface should have the Media Termination Point Required check box unchecked on the Trunk Configuration page. See Configuring a SIP Trunk on a Load Balance Point and Configuring a SIP Trunk for an Application Point for more details.

  • If you are running Cisco WebEx Meetings Server Release 2.0 or higher and using B-series blade servers, you are required to use SAN storage to fulfill the hard drive requirements. See General System Requirements for more information about SAN requirements.

Webex Productivity Tools

With Webex Productivity Tools, users can start, schedule, or join Webex Meetings from within applications, such as Microsoft® Outlook®. Users can find more information about Cisco Webex Productivity Tools at http://www.webex.com/support/productivity-tools.html.

Cisco Webex Meetings Server supports the latest versions of the Webex applications, available on the Settings > Downloads page. If users are running an older version of Webex Productivity Tools after you perform a system upgrade, they can schedule, start, and join meetings, but the latest features are not available. We recommend that you silently push the latest Webex Productivity Tools .msi for an optimal experience. (See Cisco Webex Meetings Application and Productivity Tools Compatibility Matrix.)

Users

The system supports a lifetime maximum of 400,000 user accounts. This number represents the total of both active and deactivated user accounts. This lifetime maximum number is large enough to accommodate expected growth in the user database.

Administrators cannot delete users from the system. Instead, users are deactivated. This design enables administrators to reactivate previously deactivated user accounts, even after long periods of user inactivity. The user's meetings and other content (including recordings) are restored.

Deployment Sizes For Your System

When determining the size for your system, consider how many users you expect to be using the system at any given time. For example, in a 50-user system the maximum number of users concurrently attending meetings is 50. If more than 50 users attempt to join a meeting, an error messages displays for all users who attempt to join a meeting after the maximum number of users is exceeded, and the system prevents these users from joining the meeting.

Here are some things to consider when determining the size for you system:

  • Determine the largest number of users you anticipate will join a meeting at any given time, including rare or unusual occasions.

  • You can expand the system size to a larger size at any time as long as your hardware meets or exceeds the minimum requirements for the larger size system; otherwise, you must purchase additional hardware.

  • If you plan to add High Availability (HA) or a Multi-data Center (MDC) to your system, include the additional virtual machines necessary to support the HA or MDC system, when you purchase your hardware.

Requirements for vCenter Co-residency

VMware vCenter Server or vCenter Server Appliance can reside with other virtual machines or with Cisco WebexLar Meetings Server (CWMS) virtual machines in some instances.

On a 50– or a 250–user system, VMware vCenter can reside on the same host with CWMS. However additional RAM must be installed with the Cisco UCS server. For the exact amount of RAM required, see the requirements for that system size in Minimum Hardware Requirements.

Virtual Machines In Your System

These are the virtual machines created for your system. Some functions are combined into one virtual machine for the smaller system sizes.

  • Admin—Heart node of the system. Includes the system database and provides administrative functions.

  • Media—Provides media services (audio-video function, telephony and meetings services).

    Included in the Admin virtual machine in a 50 concurrent users system.

  • Web—Provides web services (meeting list and recordings). Enables the user to schedule future meetings.

    Included in the Admin virtual machine in a 50, 250 or 800 concurrent users system.

    End users sign in to the Webex web site. Administrators sign in to the Administration web site.

  • Internet Reverse Proxy (IRP)—Provides public access, enabling users to host or attend meetings from the Internet and mobile devices. The Internet Reverse Proxy is required for your mobile workforce to attend meetings.


    Note

    Only the IRP provided with this product may be used in this system. Internet Reverse Proxies or web load balancers, supplied by other vendors, are not supported. The IRP provided with this product is optimized for handling real-time web, audio, and data-sharing traffic from external users joining meetings from the Internet.


Note

In this documentation, we use the term internal virtual machines to refer to the Admin virtual machine, and if applicable, to the Media and Web virtual machines.

The IRP is situated in the DMZ network (non-split-horizon and split-horizon network topologies) or in the internal network (all internal network toplogy).

Minimum Hardware Requirements

This section lists some of the Cisco UCS servers you can use for each size system. For specific requirements for each system, see the following topics:

Table 1. ESXi Versions and License Types

System Size

ESXi Version

vSphere License Type

50 or 250

5.0, 5.0 Update 1, 5.1, 5.5, or 6.0

Standard Edition, Enterprise Edition, Enterprise Plus Edition

800 or 2000

5.0 or 5.0 Update 1

Enterprise Plus Edition Only

5.1 to 5.52

Enterprise Edition, Enterprise Plus Edition

5.5 or higher1

Standard Edition, Enterprise Edition, Enterprise Plus Edition

6.0

Standard Edition, Enterprise Edition, Enterprise Plus Edition

2

"Or higher" references updates, not full versions. For example, 5.5 Update 1 or 5.1 Update 2.

Table 2. Host Models

Deployment Size

Example of UCS Model

50 Users

  • UCS C220 M3

  • UCS B200 M3

  • UCS C220 M4S

  • C240 M4S2

250 Users

  • UCS C220 M3

  • UCS B200 M3

  • UCS C220 M4S

  • C240 M4S2

800 Users

  • UCS C460 M2

  • UCS B440 M2

  • UCS B420 M3 (2.0 and higher)

2000 Users

  • UCS C460 M2

  • UCS B440 M2

  • UCS B420 M3 (2.0 and higher)

Co-residency with vCenter is supported with 50- and 250-user system deployments only. Co-residency with Cisco Unified Communications products on the same physical host is not supported.

You can use older models of the UCS hardware with your system, but for a better user experience use the hardware listed in the table. For example, you can use the UCS C220 M3 for a 250-user system if you already have that hardware available.

When upgrading Cisco Webex Meetings Server, you can use Cisco UCS B200 M3 blade servers with 2x local hard drives, as long as the upgraded system uses SAN storage for its virtual machines. Using SAN storage with B-series blade servers allows your system to meet the 4 hard disk drives in a RAID 5 or RAID 10 configuration requirement for Cisco Webex Meetings Server.


Note

For 800-user and 2000-user systems, we do not recommend deploying additional virtual machines on a DMZ host. This can result in increased packet loss and noticeable latency on media connections.

Resources Consumed by CWMS and the ESXi Host

Cisco Webex Meetings Server is deployed on one or more virtual machines on ESXi hosts. Both Cisco Webex Meetings Server (CWMS) and ESXi (VMware component that enables virtualization on the physical Cisco UCS Server) use CPU and memory resources, and storage space. Depending on your system size, vCenter and multiple virtual machines might run on the same Cisco UCS server.

CWMS uses resource reservation for its virtual machines to guarantee system scalability. Other VMware workloads do not take CPU and other resources away from the virtual machines. The minimum requirements for each system size includes enough resources to support:

  • Continued quality of service for CWMS at peak system usage (maximum capacity).

  • VMware ESXi.

  • VMware vCenter (when co-resident).

    For the requirements for vCenter Server, see Knowledge Base and search for "Installing vCenter Server <version> best practices," where <version> is your vCenter Server version.

  • VMware snapshots of the virtual machine (delete these as soon as possible otherwise you may experience severe performance degradation).

    Extra disk space is required for snapshots, as some snapshots may be as large as the original virtual machine. In some cases, vSphere may delete snapshots to create storage space, compromising the ability to roll back to previous snapshots.

  • Use of the Cisco UCS Server over the typical life cycle of the server.

The hardware requirements specified in the OVA file are the minimum requirements to deploy Cisco Webex Meetings Server. These requirements do not include any CPU, memory, or storage requirements for VMware vCenter or ESXi.


Caution

Co-residency, other than the configurations listed in the tables in this document, is not supported. If you disregard our system requirements, your virtual machines might not boot. The deployment of the virtual machines can stall from within the earliest product screens during the vCenter OVA deployment.

50-user System

A 50-user system is also called a micro system. (Multi-data Center (MDC) is not available for micro systems.) The diagram illustrates two versions of a 50-user deployment. (The "Redundant" virtual machines demonstrate support for High Availability (HA).)

The table lists the minimum hardware requirements for the ESXi hosts (Cisco UCS servers) in your system. The last two columns show the amount of free disk space needed for new installations of Cisco Webex Meetings Server, and for an Automatic Upgrade, using your existing Cisco UCS servers. For more information, see General System Requirements.

For information about the bandwidth requirements, see the Network Bandwidth Requirements.


Note

For IOPS information, see Advantages of Deploying a System on VMware vSphere.

Co-residency with vCenter is supported with a 50-user system deployment as configured in the following table.

Virtual Machines on ESXi Host (Cisco UCS Server)

CPU Cores

Memory (GB)

Ethernet Ports

Hard Drive Storage Requirement for New Installs

Free Hard Drive Storage Required for Automatic Upgrade3

Admin

4

 24

2 for the Admin virtual machine, including 1 if NIC teaming is used for redundancy

1 recommended for ESXi management network

1.5 TB; minimum of 7,200 RPM

1118

Admin and vCenter (co-resident)

6 (ESXi 5.0)

8 (ESXi 5.1 and above)

36

2 for the Admin virtual machine, including 1 if NIC teaming is used for redundancy

1 for vCenter

1 recommended for ESXi management network

1.5 TB; minimum of 7,200 RPM

1078

This figure is based on the assumption that you have set up vCenter with 40 GB.

Internet Reverse Proxy (IRP)

4 (ESXi 5.0)

6 (ESXi 5.1 and above)

12

2 for the IRP virtual machine, including 1 if NIC teaming is used for redundancy

1 recommended for ESXi management network

300 GB; minimum of 7,200 RPM

172
Admin and IRP (co-resident)

8

36

2 for the Admin virtual machine, including 1 if NIC teaming is used for redundancy

2 for IRP virtual machine, including 1 if NIC teaming is used for redundancy

1 recommended for ESXi management network

1.5 TB; minimum of 7,200 RPM

990
Admin and IRP and vCenter (all co-resident)

12

40

2 for the Admin virtual machine, including 1 if NIC teaming is used for redundancy

2 for IRP virtual machine, including 1 if NIC teaming is used for redundancy

1 for vCenter

1 recommended for ESXi management network

1.5 TB; minimum of 7,200 RPM

950

This figure is based on the assumption that you have set up vCenter with 40 GB.

3 If you choose Manual Upgrade where you manually deploy all the virtual machines and you are using the same hardware, you are required to have twice the free space as is required for Automatic Upgrade.

Note

If you plan to use a High Availability (HA) system, double the hardware requirements and quantities of the primary system to support both systems.

Resources Reserved by the Virtual Machines in a 50-user System

This section describes how much media the virtual machines use and is intended for those with expert knowledge of VMware. CPU resources are specified as vCPUs (cores) and MHz (CPU cycles). The VMware VMkernel uses MHz cycles to control CPU scheduling.

Memory resources are specified by maximum memory and reserved memory. Reserved memory is not shared with other virtual machines on the same physical Cisco UCS Server.

Disk resources (storage) are controlled in two separate areas. During the OVA build, the CentOS file system partition sizes determine the minimum disk size. Secondly, vCenter controls the maximum disk space available.

If you attempt to deploy a virtual machine without the minimum number of vCPUs, the OVA deployment of the virtual machine will fail. If you attempt to deploy a virtual machine without the minimum total MHz processor speed, then the virtual machine will not power on.


Important

The numbers in this table do not include resources for VMware ESXi.
Virtual Machine Type Virtual CPU (vCPU) CPU4 (MHz) Reserved Memory/Total Memory5 (GB) Disks (GB)

Admin

4

8000

12/14

418

Internet Reverse Proxy

4

8000

4/4

128
4 Number obtained by multiplying the number of physical CPUs with the speed of the CPU chip (MHz). Hyperthreading is not included in this calculation. (The physical CPU must have a clock speed of 2.4 GHz or faster.)
5 Virtual machines with media functionality have additional, non-reserved memory; Memory = Reserved/Total

250-user System

A 250-user system is also called a small system. This diagram illustrates two versions of a 250-user deployment. The "Redundant" virtual machines demonstrate support for High Availability (HA). If your system does not include HA support, only deploy the Primary system.

This diagram shows the layout of a 250-user system with two data centers that form a Multi-data Center (MDC) system with Internet Reverse Proxy (IRP) support.. The License Manager runs on only one data center.

The table lists the minimum hardware requirements for the ESXi hosts (Cisco UCS servers) in your system. The last two columns show the amount of free disk space needed for new installations of Cisco Webex Meetings Server, and for an Automatic Upgrade, using your existing Cisco UCS servers. For more information, see General System Requirements.

For information about the bandwidth requirements, see Network Bandwidth Requirements.

For IOPS information, see Advantages of Deploying a System on VMware vSphere.

Co-residency with vCenter is supported with a 250 user system deployment as configured in the following table.

Virtual Machines on ESXi Host (Cisco UCS Server)

CPU Cores

Memory (GB)

Ethernet Ports

Hard Drive Storage Requirement for New Installs

Free Hard Drive Storage Required for Automatic Upgrade6

Admin and Media 12 52

  • 2 for Admin and Media, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1.5 TB; minimum of 7200 RPM

990

(Admin and Media) and vCenter (co-resident) 16 56

  • 2 for Admin and Media, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

  • 1 for vCenter

1.5 TB; minimum of 7200 RPM

950

(This figure is based on the assumption that you have set up vCenter with 40 GB.)

Internet Reverse Proxy (IRP) 12 36

  • 2 for IRP, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

300 GB; minimum of 7200 RPM

172
6 If you choose Manual Upgrade where you manually deploy all the virtual machines and you are using the same hardware, you are required to have twice the free space as is required for Automatic Upgrade.

Note

If you plan to use a HA system, purchase the same hardware and quantities for the HA system as you did for the primary system.

Resources Reserved by the Virtual Machines in a 250-user System

This section describes how much media the virtual machines use and is intended for those with expert knowledge of VMware. CPU resources are specified as vCPUs (cores) and MHz (CPU cycles). The VMware VMkernel uses MHz cycles to control CPU scheduling.

Memory resources are specified by maximum memory and reserved memory. Reserved memory is not shared with other virtual machines on the same physical Cisco UCS Server.

Disk resources (storage) are controlled in two separate areas. During the OVA build, the CentOS file system partition sizes determine the minimum disk size. Secondly, vCenter controls the maximum disk space available.

If you attempt to deploy a virtual machine without the minimum number of vCPUs, the OVA deployment of the virtual machine will fail. If you attempt to deploy a virtual machine without the minimum total MHz processor speed, then the virtual machine will not power on.


Important

The numbers in this table do not include resources for VMware ESXi.
Virtual Machine Type Virtual CPU (vCPU) CPU7 (MHz) Reserved Memory/Total Memory8 (GB) Disks (GB)

Admin

4

8000

16/16

418

Media

8

16,480

13/23

128

Internet Reverse Proxy

8

16,480

6/6

128
7 Number obtained by multiplying the number of physical CPUs with the speed of the CPU chip (MHz). Hyperthreading is not included in this calculation. (The physical CPU must have a clock speed of 2.4 GHz or faster.)
8 Virtual machines with media functionality have additional, non-reserved memory; Memory = Reserved/Total

800-user System

An 800-user system is also called a medium system. This diagram illustrates two versions of an 800-user deployment. The "Redundant" virtual machines demonstrate support for High Availability (HA).

This diagram shows the layout of an 800-user system with two data centers that form a Multi-data Center (MDC) system with Internet Reverse Proxy (IRP) support. The License Manager runs on only one data center.

The table lists the minimum hardware requirements for the ESXi hosts (Cisco UCS servers) in your system. The last two columns show the amount of free disk space needed for new installations of Cisco Webex Meetings Server, and for an Automatic Upgrade, using your existing Cisco UCS servers. For more information, see General System Requirements.

For more information about the bandwidth requirements, see Bandwidth on Cisco Webex Meetings Server Network Interfaces.

For information about the bandwidth requirements, see the Network Bandwidth Requirements.


Note

Co-residency with vCenter is not supported with an 800-user system deployment.

For IOPS information, see Advantages of Deploying a System on VMware vSphere.

Virtual Machines on ESXi Host (Cisco UCS Server)

CPU Cores

Memory (GB)

Ethernet Ports

Hard Drive Storage Requirement for New Installs

Free Hard Drive Storage Required for Automatic Upgrade9

Admin and Media (combined) 40 80

  • 2 for Admin and Media, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1.5 TB; minimum of 10,000 RPM

900
Internet Reverse Proxy (IRP) 20 18

  • 2 for IRP, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

300 GB; minimum of 10,000 RPM

172
9 If you choose Manual Upgrade where you manually deploy all the virtual machines and you are using the same hardware, you are required to have twice the free space as is required for Automatic Upgrade.

Note

If you plan to use an HA system, purchase the same hardware requirements and quantities as the primary system.

For 800 user systems, we do not recommend deploying additional virtual machines on a DMZ host. This might result in increased packet loss and noticeable latency on media connections.

Resources Reserved by the Virtual Machines in an 800-user System

This section illustrates how much media the virtual machines use and is intended for those with expert knowledge of VMware. CPU resources are specified as vCPUs (cores) and MHz (CPU cycles). The VMware VMkernel uses MHz cycles to control CPU scheduling.

Memory resources are specified by maximum memory and reserved memory. Reserved memory is not shared with other virtual machines on the same physical Cisco UCS Server.

Disk resources (storage) are controlled in two separate areas. During the OVA build, the CentOS filesystem partition sizes determine the minimum disk size. Secondly, vCenter controls the maximum disk space available.

If you attempt to deploy a virtual machine without the minimum number of vCPUs, the OVA deployment of the virtual machine will fail. If you attempt to deploy a virtual machine without the minimum total MHz processor speed, then the virtual machine will not power on.


Important

The numbers in this table do not include resources for VMware ESXi.
Virtual Machine Type Virtual CPU (vCPU) CPU10 (MHz) Reserved Memory/Total Memory11 (GB) Disks (GB)

Admin

10

20,600

16/16

418

Media

30

60,800

14/44

128

Internet Reverse Proxy

20

41,200

10/10

128
10 Number obtained by multiplying the number of physical CPUs with the speed of the CPU chip (MHz). Hyperthreading is not included in this calculation. (The physical CPU must have a clock speed of 2.4 GHz or faster.)
11 Virtual machines with media functionality have additional, non-reserved memory; Memory = Reserved/Total

2000-user System

A 2000-user system is also described as a large system. This diagram shows a 2000-user system with High Availability (HA) and Internet Reverse Proxy (IRP) support. The HA virtual machines are shown as the Redundant virtual machines. If your system does not include HA support, only deploy the Primary system.

This diagram shows a 2000-user system deployment with two data centers that form a Multi-data Center (MDC) system with Internet Reverse Proxy (IRP) support. The License Manager runs on only one data center.


Important

We recommend that you deploy the all virtual machines shown in the diagram. By deploying different types of virtual machines on a physical server, you can better avoid a system shutdown in the event of a hardware failure. For example, placing a Media and a Web virtual machines on a single physical server is more resilient than if you place both Web virtual machines on the same physical server.

On a large system there is an exclusion from the equal load balance rule (see Load Balancing for more information), where there are SIP trunk load balancers on Media 1 and Media 2, and where Media 3 and optionally Media HA do not have load balancing. If there is a failure of both Media 1 and 2 on the primary system, all telephony service on CWMS is lost. If the system is a HA deployment, the redundancy mitigates the failure of a single virtual machine.

The table lists the minimum hardware requirements for the ESXi hosts (Cisco UCS servers) in your system. The last two columns show the amount of free disk space needed for new installations of Cisco Webex Meetings Server, and for an Automatic Upgrade, using your existing Cisco UCS servers. For more information, see General System Requirements.

For more information about the bandwidth requirements, see Network Bandwidth Requirements.

If you plan to add a HA system, those virtual machines are shown as the "redundant" virtual machines. If you do not want HA, deploy only the primary system.


Note

Co-residency with vCenter is not supported with a 2000-user system deployment.


Note

For IOPS information, see Advantages of Deploying a System on VMware vSphere.

Virtual Machines on ESXi Host (Cisco UCS Server)

CPU Cores

Memory (GB)

Ethernet Ports

Hard Drive Storage Requirement for New Installs

Free Hard Drive Storage Required for Automatic Upgrade (GB) 12

Media1 and Admin (combined)

40 80

  • 2 for Media1 and Admin, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1.5 TB; minimum of 10,000 RPM

990
Media2 and Web1 (combined) 40 80

  • 2 for Media2 and Web1, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1 TB; minimum of 10,000 RPM

768
Media3 and Web2 (combined) 40 80

  • 2 for Media3 and Web2, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1 TB; minimum of 10,000 RPM

768
Internet Reverse Proxy (IRP) 20 18

  • 2 for IRP, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

300 GB; minimum of 10,000 RPM

172
Media and Admin (combined) for HA 40 80

  • 2 for Media and Admin, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1.5 TB; minimum of 10,000 RPM

990
Web for HA 20 40

  • 2 for Web, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

1 TB; minimum of 10,000 RPM

896
IRP for HA 20 18

  • 2 for IRP, including 1 if NIC teaming is used for redundancy

  • 1 recommended for ESXi management network

300 GB; minimum of 10,000 RPM

172
12 If you choose Manual Upgrade where you manually deploy all the virtual machines and you are using the same hardware, you are required to have twice the free space as is required for Automatic Upgrade.

Note

For 2000 user systems, we do not recommend deploying additional virtual machines on a DMZ host. This might result in increased packet loss and noticeable latency on media connections.

Resources Reserved by the Virtual Machines in a 2000-user System

This section illustrates how much media the virtual machines use and is intended for those with expert knowledge of VMware. CPU resources are specified as vCPUs (cores) and MHz (CPU cycles). The VMware VMkernel uses MHz cycles to control CPU scheduling.

Memory resources are specified by maximum memory and reserved memory. Reserved memory is not shared with other virtual machines on the same physical Cisco UCS Server.

Disk resources (storage) are controlled in two separate areas. During the OVA build, the CentOS file system partition sizes determine the minimum disk size. Secondly, vCenter controls the maximum disk space available.

If you attempt to deploy a virtual machine without the minimum number of vCPUs, the OVA deployment of the virtual machine will fail. If you attempt to deploy a virtual machine without the minimum total MHz processor speed, then the virtual machine will not power on.


Important

The numbers in this table do not include resources for VMware ESXi.
Virtual Machine Type Virtual CPU (vCPU) CPU13 (MHz) Reserved Memory/Total Memory14 (GB) Disks (GB)

Admin

10

20,600

16/16

418

513

New systems must be deployed by using the OVA file .

Media

30

60,800

14/44

128

Web

10

20,600

16/16

128

Internet Reverse Proxy

20

41,200

10/10

128
13 Number obtained by multiplying the number of physical CPUs with the speed of the CPU chip (MHz). Hyperthreading is not included in this calculation. (The physical CPU must have a clock speed of 2.4 GHz or faster.)
14 Virtual machines with media functionality have additional, non-reserved memory; Memory = Reserved/Total

System Capacity Matrix

Key Points:

  • One of the basic assumptions for the information presented in this section is that there are at least two people participating in a meeting.

  • Concurrent meeting connections is defined as the number of people participating in a meeting at any given time. For example, for a 50 user system, the maximum concurrent meeting connections can be comprised of five concurrent meetings that each have a total of 10 people in the meeting (for example, one host and nine participants).

  • After the maximum number of meeting participants is reached for any point in time, the system does not allow other users to start or join meetings. Of those maximum number of meeting participants (2000, 800, 250 or 50 people), only half of the participants can use video. Video is defined as sending or receiving, meaning users might be using their Webex webcam video or the video file share option which allows users to share a video.

  • Desktop sharing is not considered video. This means with a 250 user system, 250 people can be sharing their desktops during meetings at any given time.

  • The addition of High Availability or Multi-data Center does not increase the capacity of the system to hold meetings; an 800-user system is still an 800-user system.

The numbers in the table below represent the design capacity for the Cisco Webex Meetings Server system. Operating the system at a capacity higher than these specifications can result in a degraded user experience and may result in system instability. Cisco reserves the right to enforce capacity limits at these levels.


Note
These values in the following table remain the same regardless of whether your system is a single data center or a multi-data center system.
Table 3. System Capacity Matrix

System Capacity

2000 user system

800 user system

250 user system

50 user system

Notes

Maximum Concurrent Meeting Connections (Audio, Video, and Web users)

2000

800

250

50

The number of people participating in concurrent meetings at any given time.

Maximum Simultaneous Audio Connections (Teleconference Phone Calls and Voice Connection Using Computer From Meeting Clients)

2000

800

250

50

The system capacity remains the same as shown on the left, regardless of what combination of the following features are used:

  • G.711, G.722, G.729 audio codecs

  • IPv4 or IPv6 teleconferencing

  • TLS/SRTP audio encryption

Maximum Concurrent Video and Video File Sharing Users

1000 15

400

125

25

These numbers show the maximum number of concurrent meeting connections (or participants) allowed to use video sharing at the same time. When the number of users with video sharing in concurrent meetings reaches this limit, then the remaining users invited to the concurrent meetings can join the meetings, but their video windows are dimmed.

Note 

If one participant in a meeting uses video, then all other users in the same meeting are counted as video users, even if they are not using video themselves.

Note 

Desktop sharing is not considered video.

Maximum Participants in One Meeting

500

500

250

50

These numbers show the maximum number of participants who can attend a meeting.

Maximum Meetings That Can be Recorded Simultaneously

100

40

13

3

This is the total number of meetings that can use the Recording feature at one time.

Maximum Concurrent Recording Playback Sessions

500

200

63

12

This is the total number of recording playback sessions that can occur simultaneously. This refers to recordings that are saved on your storage system and does not include recordings that are downloaded to users' desktops.

Note 

These playback sessions are not included in the concurrent meeting connections on the system.

Maximum Number of User Profiles

400,000

400,000

400,000

400,000

This number includes active and deactivated users.

Maximum Concurrent Meetings

1000

400

125

25

The number of separate meetings that can be active concurrently.

Maximum Call Rate (calls/per second)

20

8

3

1

This is the average number of users who can dial into a meeting during a one second time period. After the system reaches this number, the next few users to dial into the meeting might experience an additional few seconds wait before connecting to the meeting.

Maximum Concurrent Sign-in

20 people per second

8 people per second

3 people per second

1 person per second

This is the average number of users who can simultaneously sign in to your Webex site during a one second time period. After the system reaches this number, the next few users to sign in to the Webex site might experience an additional few seconds wait before they can join a meeting.

Maximum Aggregate Bandwidth Utilization

5 Gbps

2 Gbps

625 Mbps

125 Mbps

Using our test system at its maximum bandwidth, this is the maximum bandwidth the test system could handle. For more information about bandwidth utilization see the Network Bandwidth Requirements section in the Networking Topology for Your System chapter of the Planning Guide. You can also refer to the Webex Network Bandwidth White Paper.

15 800 in Federal Information Processing Standards (FIPS environments)

Tip

The maximum length of a meeting is 24 hours for all size user system deployments.


Note

When considering an upgrade, plan for the increased size of the data stores, as the original system and the upgraded system share data stores until testing of the upgraded system is complete and the original system is removed.

For information about network bandwidth requirements for the various size user systems, see the "Network Bandwidth Requirements" section in the Network Topology For Your System chapter in this book.

Supported Upgrade Paths

This release of Cisco WebEx Meetings Server supports upgrades from release 1.x to 2.8. The following points apply:

  • An upgrade is defined as a replacement of the system to deploy major modifications that we made to the system.

  • An update is defined as an incremental modification of the system. Updates deploy fixes and minor improvements.

  • An update retains all data from the original system. An upgrade retains all data from the original system, except for the logs.

  • When upgrading, you cannot skip a major version of the software and go directly to a companion maintenance release (MR).

    For example, to upgrade from 1.5MR5 to a 2.8MR, upgrade from 1.5MR5 to 2.8 and then update to the 2.8MR.


Note

All updates require downtime. For Multi-data centers, you update both data centers simultaneously.


Caution

Do not click Restart for one data center until the update for the other is complete, and both display the Restart button. When you update from Release 2.5MR6 or later to 2.8, restarting one data center before the update is complete for the other breaks replication.

Use the following table to determine your upgrade path to Cisco WebEx Meetings Server Release 2.8.

Installed Release

To Release

Path

1.0 to 1.1

2.8

  1. Update to 1.5.

  2. Update to 1.5MR5 Patch 2 or later.

  3. Upgrade to 2.8.

1.5 to 1.5MR4

2.8

  1. Update to 1.5MR5 Patch 2 or later.

  2. Upgrade to 2.8.

1.5 MR5

2.8

  1. Update to 1.5MR5 Patch 2 or later.

  2. Upgrade to 2.8.

1.5 MR5 Patch 2 or later

2.8

Upgrade to 2.8.

2.0 to 2.0MR8

2.8

  1. Update to 2.0MR9.

  2. Update to 2.8.

2.0MR9 or later

2.8

Update to 2.8.

2.5 to 2.5MR5

2.8

  1. Update to 2.5MR6.

  2. Update to 2.8.

2.5MR6

2.8

Update to 2.8.

2.6 to 2.6MR2

2.8

  1. Update to 2.6MR3

  2. Update to 2.8

2.6MR3 or later

2.8

Update to 2.8.

2.7 or any 2.7MR

2.8

Update to 2.8.

2.8 or any 2.8MR

Any 2.8MR

Update to 2.8MR.

Important

You cannot change the audio encryption type (Audio Encrypted -AE/Audio Unencrypted -AU) for the system, during an upgrade or during an update. After deployment, the only way to change a system from one type of audio encryption to the other is to deploy a new system.

For more information, see the following documents:

Was this Document Helpful?

FeedbackFeedback

Contact Cisco

About Us
Contact Us
Work with Us