VMware Cloud Director Availability™ 4.1 | 26 NOV 2020 | Build 17222554 (Product version: 4.1.0.1660207-48e5972493)

Check for additions and updates to these release notes.

What's in the Release Notes

The release notes cover the following topics:

What's New

VMware Cloud Director Availability 4.1 introduces the following new features:

  • Event notifications for tenants: Tenants can now receive replication events notifications by email and as VMware Cloud Director event notifications.
  • Simplified initial configuration: The initial configuration of the Cloud Service now includes all necessary steps to complete the configuration of VMware Cloud Director Availability by using a single interface, instead of separately configuring the Manager Service, the Replicator Service instances, and the Tunnel Service.
  • Improved notifications for providers: The service providers can now receive notifications by email and have more options to control the received event notification types.
  • Separate migrations and protections usage: The service provider now separately enables incoming and outgoing migrations and protections in the replication policies. In previous versions, both replication types are controlled together, as incoming replications or outgoing replications.
  • Multisite Cloud Director authentication: You can now use your external identity provider to authenticate to the remote site and manage geographically distributed installations as single entities.
  • Back up the Cloud appliances and Back up the on-premises appliance: You can now easily back up all VMware Cloud Director Availability appliances in the site. Now you can also restore each appliance by using a single backup file, see Restore the Cloud appliances and Restore the on-premises appliance.
  • Create replications for encrypted virtual machines: VMware Cloud Director Availability now supports replicating virtual machines with an encryption-enabled storage policy in VMware Cloud Director, from one cloud site to another cloud site, providing end-to-end encryption. To replicate encrypted virtual machines, verify that vCenter Server 6.7 U3 or later, VMware Cloud Director 10.1 or later, and VMware Cloud Director Availability 4.1 are installed in both the source cloud site and in the destination cloud sites.

Upgrade

VMware Cloud Director Availability 4.1 supports an in-place upgrade from versions 4.0.0.2 or 4.0.1.1. To upgrade from version 3.x, first upgrade to version 4.0.0.2. Then you can upgrade to version 4.1. 

Upgrading to VMware Cloud Director 4.1

Current Version 3.x

Current Version 4.0.x
Upgrading in the Cloud Site

First upgrade to version 4.0, by following the legacy Upgrading to version 4.0 in the Cloud procedures.

Upgrade to VMware Cloud Director 4.1 by following the updated Upgrading in the Cloud procedures.
Upgrading in the On-Premises Site First upgrade to version 4.0, by following the legacy Upgrading to version 4.0 On-Premises procedures.  Upgrade to VMware Cloud Director 4.1 by following the updated Upgrading On-Premises procedures.

Configuration Maximums

For the tested uptime, concurrency, and scale limits, see VMware Configuration Maximums.

Caveats and Limitations

For interoperability between paired sites with different VMware Cloud Director Availability versions, see Managing Connections Between Cloud Sites.

For interoperability between VMware Cloud Director Availability and other VMware products, see VMware Product Interoperability Matrices

Note: The VMware Cloud Director Availability vSphere Client Plug-In requires vSphere Client support. You can use the VMware Cloud Director Availability Portal if your vSphere environment does not support vSphere Client.

Supported Browsers

VMware Cloud Director Availability 4.1 supports the following browsers:

  • Google Chrome 84 and later
  • Microsoft Edge Chromium 84 and later
  • Mozilla Firefox 70 and later
  • Safari 14 and later

Resolved Issues

  • Performing test failover, failover, or migration from an on-premises site to a cloud site with VMware Cloud Director 10.2 shows task errors and the NICs are not connected to the correct networks

    With VMware Cloud Director 10.2 in the destination cloud site, the test failover, failover, or migration from an on-premises site succeeds but in the recent tasks you can see: 'Failed to connect NICs for vm" and "Invalid network parameter: Unknown IP Addressing Mode "DHCP" for vNIC X on VM Y connected to network "none".

  • Cannot use multisite authentication for the VMware Cloud Director Availability Plug-in after navigating to remote VMware Cloud Director

    When multisite authentication is enabled, for example source@VCD1 is associated with destination@VCD2. After you log in to the source VMware Cloud Director and switch to the destination VMware Cloud Director, when you select the VMware Cloud Director Availability Plug-in in the destination VMware Cloud Director, a login prompt appears.

  • Cannot create new replications by using the root user account with many OrgVDCs

    After logging in as the root user, in the New Replication wizard, the drop down menu for Source/Target organization is too long and is displayed outside of the window. From the list of available Virtual Data Centers, as the request URL is too long, you cannot select a Virtual Data Center, and cannot continue creating the replication.

  • Cannot create a replication when the destination storage is vVols

    The capacity of the new independent disk stored on vVols in kilobytes must be divisible by 1024. Attempting to create a replication when the destination storage is vVols fails and in the vpxd log file in vCenter Server you see the error message The requested size is not a multiple of 1MB.

  • Cannot resize an independent disk for an existing replication when the destination storage is vVols and the storage report for this replication is incorrect

    The capacity of the new independent disk stored on vVols in kilobytes must be divisible by 1024. Attempting to resize an independent disk for an existing replication when the destination storage is vVols fails and in the vpxd log file in vCenter Server you see the error message The requested size is not a multiple of 1MB.

    As a result, the storage consumption for this replication is reported incorrectly.

  • Manually configured static routes do not route the network traffic as expected

    In the management interface or in the command-line interface you see the manually configured static routes, while they do not route the network traffic as expected.

Known Issues

  • Regenerating or replacing the SSL certificate of the Replicator Service in the source site blocks the replication traffic for encrypted virtual machines

    When replicating an encrypted virtual machine, the ESXi server hosting the encrypted virtual machine establishes trust with the source Replicator Service. After the SSL certificate of the Replicator Service changes, the trust is not automatically re-established.

    Workaround: Reconfigure the replications with encrypted virtual machines to re-establish the trust.

  • Virtual machines using Virtual Trusted Platform Module prevent failover, test failover, and migration

    Attempting replication workflows for virtual machines with Virtual Trusted Platform Module (vTPM) shows the following error: Unexpected vCloud Director error. The following error was encountered while processing the OVF file you provided: Unknown device: 'Virtual TPM'

    VMware Cloud Director versions earlier than 10.2.1 do not support importing virtual machines with vTPM devices and VMware Cloud Director Availability cannot fail over, test failover, or migrate such virtual machines.

    Workaround: Upgrade to VMware Cloud Director 10.2.1.

  • VDC group networks are not visible in the network settings of vApp replications and in the recovery wizards

    VMware Cloud Director 10.2 introduces VDC groups. All of the new networks types are not listed when configuring the network settings of a vApp replication, or in the failover, test failover, or migration wizards. This issue affects all VMware Cloud Director Availability 4.x versions.

    Workaround: Configure the networks after performing a failover, a test failover, or a migration.

  • Imported networks part of VDC groups prevent failover, test failover, and migration

    VMware Cloud Director 10.2 introduces VDC groups. One of the new network types - imported networks, prevents imports from vSphere, preventing all VMware Cloud Director Availability recovery workflows. This issue occurs regardless of whether the network is in use and affects all VMware Cloud Director Availability 4.x versions.

    Workaround:

    • Before performing a failover, a test failover, or a migration, remove all imported networks from the relevant VDC group.
    • Alternatively, upgrade to VMware Cloud Director 10.2.1.
  • VMware Cloud Director 10.2 does not show the VMware Cloud Director Availability events

    In VMware Cloud Director 10.2, you cannot view the VMware Cloud Director Availability events. In the VMware Cloud Director Provider portal or in the VMware Cloud Director Tenant portal, after you navigate to the Events page, the progress does not finish spinning and the events do not show in the table. Also, in the console you see Cannot read property 'id' of null.

    Workaround:

    • Use the email delivery channel and the syslog delivery channel to view the VMware Cloud Director events.
    • Alternatively, upgrade to VMware Cloud Director 10.2.1.
  • VMware Cloud Director 10.2 shows Loading instead of the VMware Cloud Director Availability context menu for virtual machines and vApps

    In VMware Cloud Director 10.2, when you right-click a virtual machine or a vApp, the context menu for VMware Cloud Director Availability remains showing Loading... and you cannot use the VMware Cloud Director Availability context menu options.

    Workaround: 

    • To perform the menu actions, use the VMware Cloud Director Availability Plug-in. Click the More menu and select Availability (siteName).
    • Alternatively, to perform the same actions, you can open the VMware Cloud Director Availability Portal.
    • Alternatively, upgrade to VMware Cloud Director 10.2.1.
  • In the OpenApi client, GET /backups/{backupId} and GET /diagnostics/bundles/{bundleId} are listed only for documentation purposes

    The two APIs use the standard authentication method, by using a custom header "X-VCAV-Auth" and cannot download a file as they cannot pass the required custom headers when opening a new tab in the Web browser.

  • In VMware vCloud Usage Meter 3.6.1, not accepting the Cloud Service certificate leads to too many requests to the Cloud Service

    Replacing or updating the SSL certificate of the Cloud Service or re-deploying it while actively metered by vCloud Usage Meter without accepting the updated certificate leads to performance issues.
    In the vCloud Usage Meter interface, if you do not accept the updated certificate of the registered Cloud Service instance, the Cloud Service receives requests too frequently by vCloud Usage Meter.

    Workaround:

    • In the vCloud Usage Meter interface, re-register the Cloud Service and accept its SSL certificate.
    • Upgrade to vCloud Usage Meter 4.2
  • Cannot set a limit of the number of configured replications when incoming protections are disabled

    When creating or editing a replication policy with incoming protections disabled, you cannot save the value of the Limit of the number of configured replications. With incoming protections enabled, this issue does not apply.

    For example, in the Default Policy the incoming protections by default are disabled, and to configure the value for this maxincoming limit, use the OpenApi Client.

    1. In the left pane under Configuration, click OpenApi Client.
    2. In the OpenApi Client page, select post /backups > /policies > put /{id}.
    3. Under Update the policy, enter the following required fields and click Put.
    policyId: default
    displayName: Default Policy
    minRpo: 15
    maxInstances: 24
    maxPinnedInstances: 0
    maxIncoming: 0
    allowIncomingProtections: false
    allowOutgoingProtections: false
    allowIncomingMigrations: true
    allowOutgoingMigrations: true
    onpremDataCap: 0
    allowCustomSlaSettings: true
    allowNotificationConfiguration: true

    The maxIncoming value controls how many incoming replications organizations can have. A negative value represents an unlimited number of replications. With 0, you cannot use the organization as a destination. A positive value represents the upper bound. When both allowIncomingProtections and allowIncomingMigrations are false, this value is ignored.

  • Cannot assign a policy to an organization or sync with cloud by using the root user account with hundreds OrgVDCs

    If you have VMware Cloud Director 10.1.1 with hundreds of organizations, in multiple cases, VMware Cloud Director Availability shows errors, caused by the VMware Cloud Director response: BadMessageException: 500: Response header too large

    For example: 

    • After you log in as the root user, in the Assign Policy window after selecting a policy and clicking Sync with Cloud, you see Operation aborted due to an unexpected error.
    • Assigning any policy to an organization shows the same error.
    • Each hour, a scheduled task checks for organizations without an assigned policy and assigns the Default policy to them.
    • Updating any policy synchronizes the Default policy and increases the organization count in the Default policy view in time or after any policy is updated. Any organization without policy gets assigned the Default policy.
    • Installing the plug-in in VMware Cloud Director for eligible tenants: when connecting a Cloud Director instance with that many organizations to the Cloud Service during the plug-in install shows the error FailedToPublishPluginException: Unable to publish the vCloud Availability plugin to the eligible tenants. To install the plug-in, in the Cloud Service management interface, under Settings > Service endpoints > VMware Cloud Director Address click Edit, then Apply.
    • Similarly, under Settings clicking Remove plugin shows the error: FailedToUnpublishPluginException: Could not unpublish the vCloud Availability plugin for one or more tenants. The issue is the same and triggers in case any organization is left with the plug-in installed.

    Workaround: Upgrade to VMware Cloud Director 10.2.

  • Cannot create new replications by using the VMware Cloud Director Availability vSphere Client Plug-In from 4.0.x on-premises sites when paired with a 4.1 cloud site

    After pairing an on-premises site running version 4.0.x with a cloud site running version 4.1, you cannot create new migrations or new protections, both incoming or outgoing, from on-premises to cloud or from cloud to on-premises.

    When you attempt to create a replication by using the VMware Cloud Director Availability vSphere Client Plug-In, you cannot complete the wizard and you see an error message that the action is not permitted by the assigned policy. For example,

    Replication policy 'XYZ' does not allow incoming replications to be configured.

    Workaround:

    • To avoid upgrading, in the VMware Cloud Director Availability On-Premises Appliance select Allow Access from Cloud. This setting allows you to create replications by using the cloud site instead of the VMware Cloud Director Availability vSphere Client Plug-In.
    • Alternatively, upgrade the VMware Cloud Director Availability On-Premises Appliance to version 4.0.1.2 or alternatively, to version 4.1.
    • Alternatively, upgrade the cloud site to version 4.1.1.
    • After upgrading, re-register the VMware Cloud Director Availability On-Premises Appliance with the vCenter Server Lookup Service. 
  • Attempting to assign a policy when multiple pages of organizations are listed intermittently fails or replaces the existing policy assignments with the Default Policy

    In the Assign Policy dialog box, selecting organizations and clicking Assign, with more than one page of organizations listed, fails with Cannot read property 'id' of undefined or organizations listed on unvisited pages that were assigned with this policy become assigned with the Default Policy instead.

    Workaround: To ensure the correct assignment of the policy, in the Assign Policy dialog box all existing organizations must be listed on one page. When assigning a policy to the listed organizations, set the page size to the maximum of 100 results and do not limit the listed organizations by using a filter. 

    Alternatively, or with over 100 organizations, instead of assigning the policy to the organizations, assign the organizations to the policy:

    1. In the Policies screen, on the top right click Organizations.
    2. Select the organizations to assign to the policy and click Assign.
    3. In the Assign Policy dialog box, select a policy to assign to the previously selected organizations.
  • Extending the session with System administrator user cannot start outgoing replications from 4.0.x cloud sites to 4.1 or 4.2 cloud sites

    In a cloud site that is running VMware Cloud Director Availability 4.0.0.2 or VMware Cloud Director Availability 4.0.1.2, if you extend your session to a remote cloud site that is running VMware Cloud Director Availability 4.1 or VMware Cloud Director Availability 4.2 by using a System administrator user, you cannot create outgoing replications to the later site.

    Workaround: 

    • Extend your session and authenticate as a tenant user to the remote cloud site, that is running version 4.1 or 4.2. Then you can create the outgoing replication from the cloud site that is running version 4.0.x.
    • Alternatively, create the replication from the other site that is running version 4.1 or 4.2, as incoming from the cloud site that is running version 4.0.x.
    • Alternatively, upgrade the cloud site running version 4.0.x to version 4.1 or to version 4.2.
  • If configuring a replication fails during independent disk creation, this placeholder disk is not cleaned up from VMware Cloud Director

    The placeholder disk leftover is not cleaned up from VMware Cloud Director when configuring a replication fails at the creation of an independent disk. If, for example, the size of the placeholder named disk is larger than any datastore in the target storage policy can accommodate, you see a message Operation aborted due to an unexpected error. This placeholder disk, while consuming storage is not visible by using VMware Cloud Director query APIs and the independent disk in an error state is visible when browsing the destination OrgVDC.

    Workaround: In VMware Cloud Director, manually remove the placeholder independent disk that is in error state.

    To alleviate the issue, upgrade to version 4.1.1 or to version 4.2. 

  • Changing the SLA profile does not update the retention policy settings

    After changing the SLA profile from either Silver or Gold to an SLA profile that only preserves the latest instance, for example, Bronze, the retention policy settings do not update with the defined in the SLA profile.

    Workaround: Upgrade to VMware Cloud Director Availability 4.2.1.

  • The vSphere Client Plug-in of the on-premises appliance is incompatible with vCenter Server 7.0 Update 3

    After registering VMware Cloud Director Availability On-Premises Appliance 4.0.x or 4.1.0 with vCenter Server 7.0 Update 3, the VMware Cloud Director Availability vSphere Client Plug-in is not operational. In vSphere, clicking the Availability plug-in shows a Service unavailable page.

    Workaround: Upgrade to VMware Cloud Director Availability 4.1.1 or later.

  • Cannot register the plug-in for versions 4.0.x and 4.1.x in VMware Cloud Director 10.3 or later 

    When using VMware Cloud Director 10.3 or later versions, you cannot register the VMware Cloud Director Availability plug-in for versions 4.0.x and 4.1.x.

    Workaround: Upgrade to VMware Cloud Director Availability 4.2 or later.

  • Updating the independent disk size loses its assigned storage policy and is assigned with the default storage policy of the datastore

    When updating the disk size of an independent disk, the disk gets recreated but loses its assigned storage policy. VMware Cloud Director then assigns the default storage policy of the datastore to the resized disk.

    Workaround: Upgrade to VMware Cloud Director Availability 4.3.

  • Performing test failover, failover, or migrate might result in the replication failing and errors importing this virtual machine for VMware Cloud Director earlier than 10.3

    For versions of VMware Cloud Director earlier than 10.3, when performing a test failover, a failover, or a migrate, the replication might fail and show an error that VMware Cloud Director cannot find the virtual machine in the inventory in order to import it. You see the following error for this virtual machine: Trying to import a VM [id] with moref vm-id, but it is already managed by vCD.

    VMware Cloud Director 10.3 and later do not exhibit the issue.

    Workaround:

    • For a failover:
      • If the source virtual machine is still running, stop the replication, start it again, synchronize it, and failover the virtual machine.  
      • If the source virtual machine is not available, the failover cannot complete because of the issue in Cloud Director to import the virtual machine from the inventory. In the destination site Cloud Director attempt to find the virtual machine in the inventory by looking for its name and manually fix the name, network settings, and other properties of the virtual machine. However, the failover or migrate operations cannot complete and cannot be performed even by retrying this virtual machine and you can only delete this virtual machine replication.
    • For a migrate, you can stop the replication, start it again, synchronize it, and failover the virtual machine.
    • For a test-failover: perform a test clean-up first.  
      • If the cleanup fails to clean up the replication, delete the replication, start over the replication, then test-failover it.   
      • If the cleanup succeeds to clean the replication  - it might only succeed in cleaning it from the vCenter Server since the virtual machine would be auto-imported in Cloud Director but might not be cleaned from it, leaving a leaked virtual machine in Cloud Director. However, attempt performing a failover, which might also fail if auto-import is not fixed first.
check-circle-line exclamation-circle-line close-line
Scroll to top icon