Symptoms
The release task performed according to this normal procedure doesn't work so the releasing state of the task is stuck.
Through UI Panel (web):
Through CLI (it remains as ACTIVE):
CSID STATUS SPACE AVAIL REPLICAS UNIQUE IOWAIT IOLAT(ms) QDEPTH HOST
1027 active 1.79TB 1.70TB 20 0 0% 0/0 0.0 management.1e2e2cf976d44ad4.nodes.svc.vstoragedomain
1029 active 1.79TB 1.66TB 163 143 0% 0/0 0.0 management.1e2e2cf976d44ad4.nodes.svc.vstoragedomain
# vinfra node disk list | grep 1027
+--------------------------------------+--------+------+------------+-------------+----------+----------+---------------+------------+----------------+
| id | device | type | role | disk_status | used | size | physical_size | service_id | service_status |
+--------------------------------------+--------+------+------------+-------------+----------+----------+---------------+------------+----------------+
| E0D8CDD5-E3B5-472A-9A5D-99B82AE37A6D | sdd | ssd | cs | ok | 619.8MiB | 1.8TiB | 1.8TiB | 1027 | active |
Cause
Other actions were done during releasing task that causes releasing task of a Chunk Server to be stuck.
For example, you start the release task and after that without waiting until that task finish you start another different action for the same CS such as deletion or cancel many times.
Resolution
As a first step you could try updating the CS in the vstorage by using:
# vstorage -c <cluster_name> rm-cs -U 1027
If the previous one doesn't work, you can proceed with this procedure:
1) Verify if the disk is healthy to be considered in the vstorage cluster. Also, identify the mount path of the disk and whether it is well-mounted or not.
# vstorage-disks-monitor health
CS Status Node Disk Health SCSI Failures CS IOLAT(s) Disk RD_LAT(s) WR_LAT(s) SMART 5 187 188 197 198 233 Healthy
1027 active ssp-dev02 sdd 1.00 n/a 0.00 0.00 n/a n/a 0 n/a 0 0 n/a true
# vstorage -c ssp-dev2 list-services
TYPE ID ENABLED STATUS DEVICE/VOLUME GROUP DEVICE INFO PATH
CS 1027 enabled active [5653] /dev/sdd1 ATA CT2000MX500SSD1 /vstorage/80ec4a12/cs
CS 1030 enabled active [217875] /dev/sdc1 ATA CT2000MX500SSD1 /vstorage/c8a15854/cs
MDS 1 enabled active [5673] /dev/sda4 ATA SAMSUNG MZ7LH240 /vstorage/mds
# grep /vstorage/80ec4a12 /proc/mounts
/dev/sdd1 /vstorage/80ec4a12 ext4 rw,lazytime,noatime,data=ordered 0 0
2) Stop the involved CS service, umount & mount it again:
# systemctl stop vstorage-csd.ssp-dev2.1027.service
# umount /vstorage/80ec4a12
# mount /dev/sdd1 /vstorage/80ec4a12
3) Try to reset the failed service and start it:
# systemctl reset-failed vstorage-csd.ssp-dev2.1027.service
# systemctl start vstorage-csd.ssp-dev2.1027.service
4) Lastly, you could remove the CS from the vstorage cluster in order to have it clean and ready to be added again to the cluster:
# vstorage -c ssp-dev2 rm-cs 1027 --force
This command will delete the chunk server #1027 without replicating the unique chunks.
All data stored in the unique chunks on this chunk server will be lost.
Are you sure you want to proceed ? [no/yes]
yes
connected to MDS#1
Dropping CS#1027 ...
After that, you should be able to add the disk to any available Tier without issues:
Additional information
If this article doesn't solve it, you can proceed to get technical support according to the VHI documentation:
Conclusion
Remember not to execute many tasks or actions around same object or component like disk, VMs, volumes, LBs, etc simultaneously to avoid stuck results/services.