Reclaim Space which failed all the application

[hpoirier8]

Hi,

I am working on VMware vSphere 6.0 and I would like to use the "unmap" functionality at the ESX level to reclaim the unused space on the different Datastores . After that, we will be able to compact the CPG (Common provisioning Group) on our Disk array (HPe 3PAR StoreServ 8400) as we are very short in space.

But when I launch the unmap command at the ESX level, it failed a lot of applications as it seems it impairs the Disk IO,

The command I used to unmap the unused space at the ESX level is :

# esxcli storage vmfs unmap -l DS_CFG1_01

Is there someone who encounter such kind of error?

thanks

Any help would be greatly appreciate...

Thanks a lot in advance

Regards

hubert

[continuum]

> it failed a lot of applications as it seems it impairs the Disk IO,

What does that mean - did you abort the procedure to keep the VMs usable or what ?

[hpoirier8]

Hi,

thanks for this update.

We have around 20 datastores to cover our VMware needs for this Production environment.
We launched the "unmap" command only on 4 datastores and we stopped immediately when we noticed that the applications were crashing one after the other... It was becoming a real "mess" from an application point of view !!!

One think I didn't mention initially : we have two disk arrays and there is a replication (asynchronous one) at the disk array level (Array base Replication).  Maybe it could be the source of these issues, but I am really not sure, and dont understand why it could be .

Any ideas?

help would be greatly appreciated.

Rgds

[continuum]

Good catch - if possible test without that replication job running.
Makes sense to me that both tasks running at the same time could  cause corruption.

Lets try to rule that out