-
- Influencer
- Posts: 10
- Liked: 1 time
- Joined: Mar 02, 2017 8:50 am
- Full Name: Matthew McAtamney-Greenwood
- Contact:
Backup Copy Job [Feature Request]
I am using the latest release of Veeam B&R 9.5 (Update 1).
I have configured a Backup Copy Job to copy all of the various backups to an external set of rotated drives.
When viewing the copy status in the desktop client I can see that the copy job is idle (not currently running/processing any VM backups).
However, if I drill down into the copy job I can find the last VMs to be processed are still in the "Processing" state, even though their details say that there was nothing to process for those VMs.
If I then attempt to disable the copy job (because I need to change the rotated drive), The job then issues a warning that I terminated the copy job.
This is not true as it's state was idle and all previous processing was complete.
What I expect to happen when a copy interval completes is that the complete job is marked as a success and then goes into the idle state.
If disabling the job when in the idle state, this should not issue warnings about a process being terminated (as there is not process working (Because it is idle).
As I work for a software house, we have an internal policy where all warnings are treated as errors.
Currently this false warning is causing a lot of internal support requests and email notifications.
Please can this be updated to work as expected?
I have configured a Backup Copy Job to copy all of the various backups to an external set of rotated drives.
When viewing the copy status in the desktop client I can see that the copy job is idle (not currently running/processing any VM backups).
However, if I drill down into the copy job I can find the last VMs to be processed are still in the "Processing" state, even though their details say that there was nothing to process for those VMs.
If I then attempt to disable the copy job (because I need to change the rotated drive), The job then issues a warning that I terminated the copy job.
This is not true as it's state was idle and all previous processing was complete.
What I expect to happen when a copy interval completes is that the complete job is marked as a success and then goes into the idle state.
If disabling the job when in the idle state, this should not issue warnings about a process being terminated (as there is not process working (Because it is idle).
As I work for a software house, we have an internal policy where all warnings are treated as errors.
Currently this false warning is causing a lot of internal support requests and email notifications.
Please can this be updated to work as expected?
-
- Product Manager
- Posts: 20271
- Liked: 2252 times
- Joined: Oct 26, 2012 3:28 pm
- Full Name: Vladimir Eremin
- Contact:
Re: Backup Copy Job [Feature Request]
This should only happen if during a sync interval a backup copy job hasn't copied anything. If you disable a job that has already archived a restore point within current interval, no warnings will be generated. Thanks.The job then issues a warning that I terminated the copy job.
-
- Influencer
- Posts: 10
- Liked: 1 time
- Joined: Mar 02, 2017 8:50 am
- Full Name: Matthew McAtamney-Greenwood
- Contact:
Re: Backup Copy Job [Feature Request]
In our environment this is sometimes to be expected.
Most of the backups are for templates and VMs that are set to revert to a snapshot on shutdown.
Therefore a Backup Copy interval starts, determines there is nothing changed, then goes idle until the next interval.
This should be equivalent to a completed run and should not cause a warning when disabling as there is actually nothing in-flight.
At the very least this should be an option within the Backup Copy Job configuration?
To set it so that each interval is considered a run rather than, if nothing happened then just sit and wait until something does happen.
Also I find this is also the behaviour when regardless of how many Backup Jobs are part of the Copy Job, and regardless of how many VMs are in each Backup Job.
If the last one had no changes (regardless of whether any other VMs had changes) it also sits and hangs in the idle state where disabling will fire the same warning.
Most of the backups are for templates and VMs that are set to revert to a snapshot on shutdown.
Therefore a Backup Copy interval starts, determines there is nothing changed, then goes idle until the next interval.
This should be equivalent to a completed run and should not cause a warning when disabling as there is actually nothing in-flight.
At the very least this should be an option within the Backup Copy Job configuration?
To set it so that each interval is considered a run rather than, if nothing happened then just sit and wait until something does happen.
Also I find this is also the behaviour when regardless of how many Backup Jobs are part of the Copy Job, and regardless of how many VMs are in each Backup Job.
If the last one had no changes (regardless of whether any other VMs had changes) it also sits and hangs in the idle state where disabling will fire the same warning.
-
- Product Manager
- Posts: 20271
- Liked: 2252 times
- Joined: Oct 26, 2012 3:28 pm
- Full Name: Vladimir Eremin
- Contact:
Re: Backup Copy Job [Feature Request]
I don't believe so, since there is no guarantee that no new restore points to copy will appear during current interval.This should be equivalent to a completed run and should not cause a warning when disabling as there is actually nothing in-flight.
Therefore, when you disable a backup copy job, you basically terminate it before it does its duty. And that's what the said warning suggests.
Anyway, we do appreciate it feedback; thank you again for raising this issue. We might consider changing existing warning pattern down the road.
-
- Influencer
- Posts: 10
- Liked: 1 time
- Joined: Mar 02, 2017 8:50 am
- Full Name: Matthew McAtamney-Greenwood
- Contact:
Re: Backup Copy Job [Feature Request]
So maybe I have the wrong configuration.
I have my backup copy job set to 1 day interval (so the files roll-over on a daily basis).
During this the copy job seems to activate every 2 mins approx to check for changes.
Am I wrong in thinking that these checks for changes should not interfere and cause this warning?
Considering my copy job is just set to copy backups that have already been performed, does it actually need to sit and poll for changes?
Should this not just wait for each backup job to complete (I'm sure there is an event for this as I can setup SNMP ping when job completes).
Even with the polling, as it currently appears to work, why does it produce the warning while it is idle?
Surely the fact it is idle means there is nothing being processed and therefore is safe to disable?
I have my backup copy job set to 1 day interval (so the files roll-over on a daily basis).
During this the copy job seems to activate every 2 mins approx to check for changes.
Am I wrong in thinking that these checks for changes should not interfere and cause this warning?
Considering my copy job is just set to copy backups that have already been performed, does it actually need to sit and poll for changes?
Should this not just wait for each backup job to complete (I'm sure there is an event for this as I can setup SNMP ping when job completes).
Even with the polling, as it currently appears to work, why does it produce the warning while it is idle?
Surely the fact it is idle means there is nothing being processed and therefore is safe to disable?
-
- Product Manager
- Posts: 20271
- Liked: 2252 times
- Joined: Oct 26, 2012 3:28 pm
- Full Name: Vladimir Eremin
- Contact:
Re: Backup Copy Job [Feature Request]
Because it has not copied anything within the current interval. So, in other words it's failed to perform its duty - create a restore point within allocated sync inteval.Even with the polling, as it currently appears to work, why does it produce the warning while it is idle?
Not exactly. The fact that backup copy job has not copied anything yet does not necessarily mean that it won't copy something before current interval finishes.Surely the fact it is idle means there is nothing being processed and therefore is safe to disable?
At what time does a source backup job start? And at what time does a sync interval for a backup copy job start?I have my backup copy job set to 1 day interval (so the files roll-over on a daily basis).
-
- Influencer
- Posts: 10
- Liked: 1 time
- Joined: Mar 02, 2017 8:50 am
- Full Name: Matthew McAtamney-Greenwood
- Contact:
Re: Backup Copy Job [Feature Request]
This I think is irrelevant. If there is nothing to update then how can it be a failure? Granted nothing completed processing, but also nothing started processing.Because it has not copied anything within the current interval. So, in other words it's failed to perform its duty.
In the case where the copy job is going to a rotated drive repository, should there not be an option to suspend while drives are being rotated?The fact that backup copy job has not copied anything yet does not necessarily mean that it won't copy something before current interval finishes.
Ultimately I am not trying to "cancel" the process, I just don't want errors if it tries to do a check while I have disconnected the drive during a rotation.
My actual backups are configured differently for each environment (users based in different timezones across the world).At what time does a source backup job start? And at what time does a sync interval for a backup copy job start?
Therefore their backups are performed at a time their users are not using the VMs (Although these VMs may still be powered on; e.g. Domain Controllers).
The backup copy job is set for copying all changes throughout the day and is set to "roll-over" at midnight UTC (so there is only a single incremental backup for each day of the week).
The rotated drives can be replaced at any point during the normal UK working day.
-
- Product Manager
- Posts: 20271
- Liked: 2252 times
- Joined: Oct 26, 2012 3:28 pm
- Full Name: Vladimir Eremin
- Contact:
Re: Backup Copy Job [Feature Request]
Firstly, how you can be sure that there is nothing to update, if the current interval is yet to expire? There is no guarantee that no new source restore points will appear.If there is nothing to update then how can it be a failure?
Secondly, it's not a failure, but rather a warning.
Disable option does exactly that. Though, by design if you disable a job within currently lasting interval, and no restore points have been copied within it, a warning will appear.In the case where the copy job is going to a rotated drive repository, should there not be an option to suspend while drives are being rotated?
Anyway, it seems that this discussion is running in circles. I think I got your feature request, so consider your voice counted.
Who is online
Users browsing this forum: srlarsen and 258 guests