"/opt/vmware/support" directory frequently becomes full within the VMware Live Recovery Appliance
search cancel

"/opt/vmware/support" directory frequently becomes full within the VMware Live Recovery Appliance

book

Article ID: 450977

calendar_today

Updated On:

Products

VMware Live Recovery

Issue/Introduction

  • Space constraints are noticed inside the "/opt/vmware/support" folder within the VMware Live Recovery Appliance
  • The directory and the partition often gets full
  • /opt/vmware/support/logs/hms directory has a list of big files

Environment

VMware Live recovery Appliance 9.0.3 and above

 

Cause

  • The hms-stderr.log file is deleted by logrotate but still referenced by the hms process.
  • The HMS process holds a massive log file that was technically "deleted" from the file system.
  • Because the process is still running, the disk space isn't actually released, causing the "ghost" file to grow indefinitely.

Resolution

This issue us fixed in Protection and Recovery 9.1.1

Workaround:

  1. Login into the VLR appliance using "admin" credentials.
  2. Change user to "root".
  3. Stop the HMS service: systemctl stop hms
  4. Navigate to /opt/vmware/support/logs/hms
  5. Verify the (deleted) files by running: lsof | grep deleted
  6. Remove all hms-stderr.log* and hms-stdout.log* files by running:
    rm hms-stderr.log*
    rm hms-stdout.log*
  7. Navigate to /etc/logrotate.d/
  8. Edit file hms-service.conf to remove "nocreate" and add "copytruncate" as per below:
    vi hms-service.conf

    /var/log/vmware/hms/hms-stdout.log
    /var/log/vmware/hms/hms-stderr.log {
      size 50M
      rotate 5
      compress
      copytruncate
      nodateext
      su root vrms
    }
  9. Verify the change
    cat /etc/logrotate.d/hms-service.conf
  10. Start the HMS service systemctl start hms
  11. Restart the logrotate service : systemctl restart logrotate