What’s new in Workload Balancing
The latest version of the Workload Balancing virtual appliance is version 8.3.0. You can download this version of the Workload Balancing virtual appliance from the XenServer Downloads page.
What’s new in 8.3.0
Released Feb 23, 2023
This update includes the following improvements:
- You can now set the alert level for Workload Balancing alerts in XenCenter by using the Management API.
This update includes changes to the WLB database. Ensure that you use the provided migration script when you update your WLB to this version. For more information about using the migration script, see Migrate data from an existing virtual appliance.
Fixed issues in 8.3.0
This update fixes the following issues:
- During the Workload Balancing maintenance window, Workload Balancing is unable to provide placement recommendations. When this situation occurs, you see the error: “4010 Pool discovery has not been completed. Using original algorithm.” The Workload Balancing maintenance window is less than 20 minutes long and by default is scheduled at midnight.
- For a Workload Balancing virtual appliance version 8.2.2 and later that doesn’t use LVM, you cannot extend the available disk space.
- Due to an unresponsive API call, Workload Balancing is sometimes blocked during pool discovery.
- In XenCenter, the date range and some timestamps shown on the Workload Balancing Pool Audit Report are incorrect.
- In XenCenter, some strings are not displaying correctly for Workload Reports.
- If the Workload Balancing virtual appliance is running for a long time, it is shut down by the operating system for consuming a lot of memory.
- The database fails to auto-restart after the Workload Balancing virtual appliance experiences an abnormal shutdown.
Known issues in 8.3.0
This update contains the following known issues:
-
If you do not use the default user name (
postgres) for the Workload Balancing PostgreSQL database, you cannot connect your pools to the Workload Balancing virtual appliance.To work around this issue, use the default user name
postgresfor the Workload Balancing database. -
After configuring Workload Balancing on a pool, if your host CPU load is low, you might encounter an
Internal errorwhen attempting to put your host into maintenance mode.To work around this issue, pause the Workload Balancing virtual appliance or disconnect it from your pools.
-
The self-signed certificate generated during the first setup of the Workload Balancing virtual appliance is only valid for the hostname or IP address, not the FQDN.
Earlier releases
This section lists features in previous releases along with their fixed issues. These earlier releases are superseded by the latest version of the Workload Balancing virtual appliance. Update to the latest version of the Workload Balancing virtual appliance when it is available.
Workload Balancing 8.2.2
Released Sep 30, 2021
This update includes changes to the WLB database. Ensure that you use the provided migration script when you update your WLB to this version. For more information about using the migration script, see Migrate data from an existing virtual appliance.
Fixed issues
This update includes fixes for the following issues:
- The Workload Balancing database can grow very fast and fill the disk.
- A race condition can sometimes cause records to be duplicated in the WLB database. When this occurs, the user might see the error: “WLB received an unknown exception”.
Workload Balancing 8.2.1
Released Sep 15, 2020
This update includes the following improvements:
-
The migration script now enables you to migrate your Workload Balancing database from the Workload Balancing virtual appliance 8.0.0 (which was provided with Citrix Hypervisor 8.0 and 8.1) to the Workload Balancing virtual appliance 8.2.1 provided with Citrix Hypervisor 8.2.
For more information about using the migration script, see Migrate data from an existing virtual appliance.
Fixed issues
This update includes fixes for the following issues:
- When multiple VMs start at the same time, Workload Balancing recommends balancing the VMs placement on all hosts in the pool evenly. However, sometimes Workload Balancing might recommend to put many VMs on the same XenServer host. This issue occurs when Workload Balancing gets late feedback from XAPI about VM placement.