Manually migration volume from pool to another in order to free space

[AlexStelmaher]

Is your feature request related to a problem? Please describe.
When created persistent volumes with mayastor storage class, volumes randomly allocated to to random pool, in this situation some/all big volumes may be allocated only to the one pool . this may cause pool running out of space.
Describe the solution you'd like
Need possibility to move specific volume/replica manually/automatically to another pool in order to free space.

Describe alternatives you've considered
Possible control on which pool volume/replica will be allocated, when creating new pvc/pv .
Additional context

`` km get pools
ID DISKS MANAGED NODE STATUS CAPACITY ALLOCATED AVAILABLE COMMITTED
pool-on-kubdockerlnx03 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:1:0?uuid=755569b5-1c8b-4541-a035-2b8d2165d3a5 true kubdockerlnx03 Online 699.3GiB 372MiB 698.9GiB 32.1GiB
pool-on-kubdockerlnx03-2 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:2:0?uuid=ba53fafe-7867-49d1-be5f-7fb5ee9ad30a true kubdockerlnx03 Online 9TiB 6.4GiB 9TiB 670.1GiB
pool-on-kubdockerlnx02 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:1:0?uuid=079e5049-bf03-46ac-8a3e-f8c8d95852d9 true kubdockerlnx02 Online 699.3GiB 48GiB 651.3GiB 110GiB
pool-on-kubdockerlnx02-2 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:2:0?uuid=8dc0f48c-9558-436f-ad69-77404954bf9a true kubdockerlnx02 Online 9TiB 6.4GiB 9TiB 620GiB
pool-on-kubdockerlnx01 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:1:0?uuid=d9eaa196-a605-435d-8825-00fc3e2d5589 true kubdockerlnx01 Online 699.3GiB 176MiB 699.1GiB 20GiB
pool-on-kubdockerlnx01-2 aio:///dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:2:0?uuid=bca0d512-c0e3-4df9-88a5-34ad8e2286aa true kubdockerlnx01 Online 9TiB 6.1TiB 2.9TiB
can see that pool-on-kubdockerlnx01-2 with capacity 9T and allocated space 6.1T and pool-on-kubdockerlnx02-2 with same capacity and allocated space 6.4G

[tiagolobocastro]

As a temporary measure you could scale the volume up and down.

[AlexStelmaher]

As a temporary measure you could scale the volume up and down.

@tiagolobocastro
Do You mean increase and decrease pvc size of volumes that belongs to the pool?

[tiagolobocastro]

No, I mean scaling the volume replicas, ex: if the current volume has 1 replica
kubectl-openebs mayastor scale volume 2
Wait for rebuild to complete (volume must be mounted by a pod)
kubectl-openebs mayastor scale volume 1

[AlexStelmaher]

@tiagolobocastro i scaled volume up and down,but after scaling down replica of volume remained on original pool. my intention it to move replica to another pool

[tiagolobocastro]

Sorry missed a step;
Before scale down you have to cordon the pool where you want to move out of.

[AlexStelmaher]

@tiagolobocastro Thanks for response . no option to cordon pool in ver 4.01 but i handled the issue by scaling volume up , io-engine-client replica destroy, and scaling volume down.

[tiagolobocastro]

ha nice wa :-)