This process restores all collections in the Arango database backup.
CAUTION: Take care when running a restore operation to avoid undesired
operational effects.
- Secure shell into the KVM master
node.
- Navigate to the
/home/labuser/platform/ansible/install directory.
- Extract the deploy services
script:
./deploy_services.sh
–x
- Run the restore process:
./arango-restore.sh
<backup location> where
<backup location> is the directory on the DM host
VM where the backup is located starting from
/halo_cluster/persistence/csdm-csdm-backup-vol-pvc-<unique K8s
generated ID>/. So, the supplied path should start with the
date.
./arango-restore.sh
2023-03-09/arango/arangobck20230209143339
- To verify the restore job:
kubectl -n csdm get jobs |
grep <backup job name> where
<backup job name> is the user-defined, unique job
name.
- To view the restore job's (POD)
log information:
kubectl get pods -A | grep -i
restore
kubectl -n csdm logs
jobs/<restore POD name> where
<restore POD name> is the manual backup's arango
restore POD name.
- Restart the VTM container
POD:
kubectl get pods -A | grep
vtm
kubectl -n csdm delete pod
<VTM POD ID> where <VTM
POD ID> is the VTM's POD ID.
Example:
labuser@vtmhalo:~$ cd platform/ansible/install
labuser@vtmhalo:~/platform/ansible/install$ ./arango-restore.sh 2023-02-09/arango/arangobck20230209143339
job.batch/backup-arango-restore-20230209193914 created
labuser@vtmhalo:~/platform/ansible/install$
labuser@vtmhalo:~/platform/ansible/install$ kubectl -n csdm get jobs backup-arango-restore-20230209193914
NAME COMPLETIONS DURATION AGE
backup-arango-restore-20230209193914 1/1 7s 68s
labuser@vtmhalo:~/platform/ansible/install$ kubectl get pods -A | grep -i restore
csdm backup-arango-restore-20230209193914-z9hxw 0/1 Completed 0 2m5s
labuser@vtmhalo:~/platform/ansible/install$
