Vras-vradmind service

Veritas VVR / VRAS service check.

Use this page when vradmin has problems, replication status cannot be checked, or the VRAS daemon shows failed.

Check service status

service vras-vradmind status

Problem example

VxVM VVR  V-5-4-0 VRAS daemon running: [  FAILED  ]
VxVM VVR  V-5-4-0 vvr_stats running: [  OK  ]

This means:

VRAS daemon is failed
vvr_stats is OK

Fix

Restart the service:

service vras-vradmind restart

Verify service

service vras-vradmind status

Expected result:

VxVM VVR  V-5-4-0 VRAS daemon running: [  OK  ]
VxVM VVR  V-5-4-0 vvr_stats running: [  OK  ]

Verify replication status

After the service is OK, check replication again:

vradmin -g DISK_GROUP repstatus RVG_NAME

Example:

vradmin -g dg_example repstatus dg_example_rvg

Good signs:

Data status:                consistent, up-to-date
Replication status:         replicating (connected)
Timestamp Information:      behind by  0h 0m 0s

If vradmin still does not work

Check cluster status:

hastatus -sum
hagrp -state
hares -state

Things to check:

Is the correct service group online?
Is a required Veritas resource offline or faulted?
Is the server/node in a healthy cluster state?
Is the related service group running on the expected node?

Useful log checks

sudo tail -n 100 /var/log/messages
sudo grep -i "vras" /var/log/messages
sudo grep -i "vradmind" /var/log/messages
journalctl -xe

Safe troubleshooting flow

1. Run service vras-vradmind status.
2. If VRAS daemon is FAILED and vvr_stats is OK, restart vras-vradmind.
3. Verify service status again.
4. Run vradmin repstatus again.
5. Confirm Data status is consistent and up-to-date.
6. If vradmin still cannot connect, check hastatus -sum.
7. If cluster resources are faulted/offline, investigate or escalate.

Commands to remember

service vras-vradmind status
service vras-vradmind restart
vradmin -g DISK_GROUP repstatus RVG_NAME
hastatus -sum
hagrp -state
hares -state

Notes

This environment uses the service command for this Veritas service.

Do not automatically replace it with systemctl unless verified.

Do not restart services blindly on production without understanding the impact.