Thanks for the response. I appologize. I should have been more clear.
I'm doing snapshots of iSCSI LUNs for VMWare. I have figured out how to create a snapshot and schedule the one snapshot. I have also figured out how to make the snapshotted LUN available in VMWare to mount and restore individual files from it. This part is excellent.
What I need to know is if it is possible *using the same snapshot volume* to create multiple snapshots back in time on a hourly, daily, weekly basis. Or do I need to create a snapshot volume for each snapshot I would like to take?
Is there a good way to estimate how much data will change when considering how big to make the snapshot volumes?
It seems that the open-e snapshot system is not designed for long term snapshots (ie keeping a yearly snapshot) but is more for temporary moments in time. Is that assumption correct?
This is great! Could you post your VMware snapshot results in the VMware section – others could benefit from this as we don't have much information with VMware other then our customers posting results on this section.
Yes you will need to take a snapshot for each Volume, you can create up to 10 but this can be a performance hit. Estimate the amount of changes you expect and set this size. Snapshot is not designed for long term, our research and development team will be working on developing more advancements to this in the future but no date is set as of yet.
Interesting. I've been doing addiational testing for this.
It seems that you can schedule multiple snapshots to occur for the same snapshot volume (ie. one at 2:00 am and one at 12:00 noon) and it takes the first snap but when the second one rolls around nothing happens. I've looked through logs for errors and I can't find it. Which one should I look for? If it's using LVM to do snapshots (which it looks like it's doing) those errors should be in the system log (ie. messages). Where can I find that?
Is this expected or should I not be able to schedule more than one snapshot for the same snap volume?
When reviewing the logs for the Snapshot and errors download the logs then go to the Task directory and critical_error logs to see if you have errors that are not presenting themselves. Could you also send your logs to support@open-e to have our engineers review them to see if there is an issue.
Also I wanted to update you that if you are using the 64 bit kernel there is a current bug, so we have blocked this when running more than one Snapshot until new fix has been completed.
If you went into the Console screen then ctrl. + alt. + t (Console tools) then selected Boot Options > System Architecture > Single or Multi proc. 64bit. But it does not sound like you went this deep into it. Send logs to support and ask them to check this issue with the Snapshot when using 2 on separate times on same volume. Could be something else is going on.
Reply With Quote
