Re-balancing replications between Hbrsrvuws
search cancel

Re-balancing replications between Hbrsrvuws

book

Article ID: 381647

calendar_today

Updated On:

Products

VMware Live Recovery VMware vSphere ESXi VMware Site Recovery Manager 8.x

Issue/Introduction

  • ESXi host frequently logs the tasks shown in the screenshot below overwhelming other tasks & this could also fill out the vCenter event tables fast if these messages keep logging at this rate. Because re-balancing of the workload occurs automatically every 30 minutes, these messages will also get logged in the interval of 30 minutes.  




Task Name : Re-balancing replications between Hbrsrvuws
Status        : Completed
Initiator      : com.vmware.vr-sa-########-####-####-####-############@vsphere.local
Target        : vrms.vmware.com
Server       : vrms.vmware.com
Related events:
Date Time    Description
08/08/2023, 2:43:45 AM    Task: Re-balancing replications between Hbrsrvuws

Environment

  • vSphere Replication 8.x
  • vSphere Replication 9.x
  • VMware Live Recovery 9.0
  • Protection and Recovery 9.1

Cause

  • vSphere Replication has undergone some major enhancements in this release. The now Enhanced vSphere Replication includes automated load balancing and scaling to achieve higher performance, and when enabled with VMware Live Recovery provides a 1 minute RPO.
  • Traditional vSphere Replication routes traffic to the replication appliance at the target site. With Enhanced vSphere Replication, replication traffic goes directly to the host at the target site, bypassing the replication appliance while supporting backward compatibility at the same time supporting a more optimized data path allowing for increased simplicity, automatic scalability and automatic load balancing.
  • This release of vSphere Replication supports both versions of the data paths, while the next release will deprecate the traditional vSphere Replication data path and replication through the appliance in favor of Enhanced vSphere Replication. Switching over is simple and easy, and non-disruptive with details included in the documentation. See How do I reconfigure existing replications to use vSphere Replication with enhanced replication capabilities.

    This is part of Enhanced vSphere Replication logging that has no bearing or impact on legacy replications. Legacy replications won't be rebalanced. Enhanced replication distributes the replications on all available hosts in the target cluster. Re-balancing of the workload occurs automatically every 30 minutes.

Resolution

  • This is a workaround to stop the replication server from logging the rebalancing messages in the ESXi host which is an informational update. 

    NOTE: Do not disable scale-out-mode, if you are using enhanced replications 

1. SSH to vSphere Replication or the VMware Live Recovery (for versions later than 9.0.3) appliance. Run the command - systemctl stop hms

2. Edit /opt/vmware/hms/conf/hms-configuration.xml and change scale-out-mode to false.

Note - For versions later than 9.0.3, the path to the configuration file is /opt/vmware/etc/hms/hms-configuration.xml

<scale-out-mode>false</scale-out-mode>
<!--
Timeout to wait before tagging hbrsrv as decommissioned due to maintenance mode.
At the moment set to 0 since replications are auto-released on hbrsrvuw when ESX host enters MM.
-->

3. Start HMS service - systemctl restart hms

What is scale-out mode ?

VMware Site Recovery now supports automated load balancing which distributes replicated VMs across resources for optimal performance. In addition, VMware Site Recovery supports automated scale-out which utilizes provisioned ESXi hosts and automatically adds them for additional replication capacity.


OR
 

  • This modification will limit the replication server from rebalancing replications on the ESXi hosts and stop the frequent loggings that show up in the tasks pane every time rebalances trigger. 

1. SSH to vSphere Replication appliance. Run the command - systemctl stop hms

2. Edit /opt/vmware/hms/conf/hms-configuration.xml and increase hbrsrv-monitor-max-rebalance-delay-sec.

Note - For versions later than 9.0.3, the path to the configuration file is /opt/vmware/etc/hms/hms-configuration.xml

<!-- do not rebalance less often then the specified number of seconds -->
<hbrsrv-monitor-max-rebalance-delay-sec>1800</hbrsrv-monitor-max-rebalance-delay-sec>

Changing this value to 86400 will set the replication rebalances to occur once in 24 hours. You can tweak this setting depending on the number of hours you want the replications to be rebalanced at. Ideally twice a day should be good. 

3. Start HMS service - systemctl restart hms

Additional Information

HBRsrvUW - Short for HBR Server User World. A process running in VMkernel is called a user world.

HMS  > HBR management system
hbrsrv > Host-based replication sever

Powered by Wolken Software