Large deployments require increasing vSphere Replication Server memory

Products

• VMware Live Recovery
• VMware vSphere ESXi

Issue/Introduction

On very large deployments, vSphere Replication Server can fail during some SRM workflows. The server restarts, but might fail again, or the recovery might fail.

 

Resolution

1. For the additional vSphere Replication server appliances in the inventory, increase the virtual machine memory size after checking the current memory consumption. To check the current memory consumption of the appliance, run the command below.

     root[]# systemctl status hbrsrv
  ●  hbrsrv.service - Host-based replication server.
     Loaded: loaded (/usr/lib/systemd/system/hbrsrv.service; disabled; vendor preset: enabled)
     Active: active (running) since Sat 2025-03-29 00:08:31 UTC; 3 weeks 2 days ago
     Main PID: 1347 (hbrsrv-bin)
      Tasks: 33 (limit: 14311)
     Memory: 625.9M (limit: 640.0M)
     CGroup: /system.slice/hbrsrv.service
             ├─1347 /usr/bin/hbrsrv-bin --daemon --pidfile /var/run/hbrsrv/hbrsrv.pid --vmodlport 8123 --lwdport 31031       --lwdsport 32032
             └─1348 /usr/bin/logger -p daemon err -t hbrsrv.stdout[1347]

MAX_MEM_SIZE = Max number megabytes of resident system memory the hbrsrv may consume. Note that this is _resident_ memory. So it is possible for the hbrsrv to consume more than this if there is available swap space. Default is 640.

2. Increase memory size from VM Edit Settings for the VRMS & the VR Add-ons

3. Increase the MAX_MEM_SIZE for the vSphere Replication Server in the VRMS and VR Add-ons:

                 a. Log into the appliance.
                 b. In the /etc/hbrsrv-init.config file change the MAX_MEM_SIZE setting from 640 to 1024 or higher depending on the limit you see for hbrsrv. Look at the command above. The value must be set in megabytes.

4. Restart the appliances after performing the steps above.