Re: Asking this in dev as there are many fixes for volumes/snapshots on 4.15.2 and 4.16.0
Hi Kristaps, With VMware, I could reproduce this issue where multiple volumes are transferred to the secondary storage. This is a VMware only issue, other hypervisors work fine. I've created a github issue to fix/track it https://github.com/apache/cloudstack/issues/5498 [https://opengraph.githubassets.com/9a7b1ad38b90cdbff8c4bd57ad5ef68e9ffe917368083c7b3d1527690c539b79/apache/cloudstack/issues/5498]<https://github.com/apache/cloudstack/issues/5498> Single volume snapshot operation exports all volumes of the VM to the secondary storage · Issue #5498 · apache/cloudstack<https://github.com/apache/cloudstack/issues/5498> ISSUE TYPE Bug Report COMPONENT NAME VMware, Volume Snapshot CLOUDSTACK VERSION 4.15.2, main CONFIGURATION OS / ENVIRONMENT SUMMARY When a VM has multiple volumes and snapshot or recurring snapshot... github.com I'll start working on this issue and update it. Thanks for letting us know. Regards, Harikrishna From: Kristaps Cudars Sent: Friday, September 17, 2021 2:40 PM To: dev@cloudstack.apache.org Subject: Asking this in dev as there are many fixes for volumes/snapshots on 4.15.2 and 4.16.0 4.15.1/Vmware When you create reoccurring volume snapshot task, instead of transferring only that volume all volumes that are associated with VM are transferred to secondary storage. Way how we find this is that we have some VM with 8~TB disks that should not be backed up and our secondary storage now has increased in size 10 times. Maybe this is fixed/resolved with different wording?
Re: Asking this in dev as there are many fixes for volumes/snapshots on 4.15.2 and 4.16.0
I can't replicate this on 4.16/KVM. Scheduled volume(!) snapshot functionality only backs up the respective volume, as expected. Regards On 2021-09-21 16:59, Riepl, Gregor (SWISS TXT) wrote: When you create reoccurring volume snapshot task, instead of transferring only that volume all volumes that are associated with VM are transferred to secondary storage. That doesn't sound right. Are you sure you're really creating a recurring volume snapshot and not a VM snapshot? I believe the VM snapshot includes all disks, but the volume snapshot should not. Related API: https://cloudstack.apache.org/api/apidocs-4.15/apis/createSnapshot.html https://cloudstack.apache.org/api/apidocs-4.15/apis/createVMSnapshot.html
Re: Asking this in dev as there are many fixes for volumes/snapshots on 4.15.2 and 4.16.0
When you create reoccurring volume snapshot task, instead of transferring only that volume all volumes that are associated with VM are transferred to secondary storage. That doesn't sound right. Are you sure you're really creating a recurring volume snapshot and not a VM snapshot? I believe the VM snapshot includes all disks, but the volume snapshot should not. Related API: https://cloudstack.apache.org/api/apidocs-4.15/apis/createSnapshot.html https://cloudstack.apache.org/api/apidocs-4.15/apis/createVMSnapshot.html
Asking this in dev as there are many fixes for volumes/snapshots on 4.15.2 and 4.16.0
4.15.1/Vmware When you create reoccurring volume snapshot task, instead of transferring only that volume all volumes that are associated with VM are transferred to secondary storage. Way how we find this is that we have some VM with 8~TB disks that should not be backed up and our secondary storage now has increased in size 10 times. Maybe this is fixed/resolved with different wording?
