Post Job Activity and Retry Job

[htwnrva]

Is there a way to stop the post-job activity from happening when I "retry job"? I'm seeing where I kick off a tape backup after the job which is ~5 hours of tape and if I retry a failed job with 1 or 2 VMs in it - it kicks off another backup job. I would prefer to only do the post-backup job during the normal scheduled run or only after the retries.

[Gostev]

Hi Ronny, right now there is no way to control script behavior in the product. But your last sentence is confusing, can you clarify when you want the script to run, and when you don't want it to run?

[htwnrva]

What i meant was:

If the job is successful or warning: proceed with the post-backup activity
If the job errors - do not proceed with the post-backup activity; Once I "Retry job" and it's successful - then proceed with the post job activity

[Gostev]

Ronny, basically your description outlines how post-job scripts works today, some other customers have even asked to change that before
I am only not sure about the very last part in your description: if it works on successful retry. I will check with devs on that.

[JacobLund]

I was just wondering. If "post job activity" is called even if the job is in retry state, then what function should be used to call a remote replication job? After veeam has finished I need to replicate the files to an hosted external storage.

[Gostev]

Please search the forum for PowerShell script that I have posted here earlier, this can be used to invoke file replication utility only after backup is completed successfully. Thanks.

[KOJGrIT]

Tried to search the script on the forum whit no luck. Can some one post a link?

[Vitaliy S.]

Kolbjørn,

I believe Anton was referring to this post, that shows how to return job status before triggering a subsequent replication job:
Veeam Backup v4 wait command (nevermind the title)

Thanks.

[Gostev]

Actually I meant this more recent script. This script launches tape backup tool once the Veeam job completes successfully. You just need the IF statement from it, as this is what queries the job status. Hope this helps.

[flavor4real]

I struggled with this every here and then too. I have a Backup Job that triggers 2 other jobs in chain. If the main job fails and retries, it usually process the complete chain again. but don;t worry, because a incremental only backs up the changes so if the job retries and there were only minimal changes on the daisy chained jobs then the data size will be low.

I'm not sure about replication because I have no replication implemented (yet).

It might be a good option to have a function which would give us the option to follow through with the remaining daisy chained jobs if a retry is in place. the reason is that I had at times to wait to have all jobs completed in order to restart the server, or do something else. Half of the time, when I stopped a job or daisy chained job, I had issues with the Veeam application itself. Either it hang, or it performed worse for a while or it even took forever to stop a job. So, I learned not (try not to) stop any jobs.

Thanks.

[KOJGrIT]

Thanks for the link

[damocue]

[merged]

Hello,

I've read the documentation and searched on forums, but I didn't find anything about to execute post-job scripts only when it finishes with succesful state. Why I miss this option? Well, I run a copy script after job finishes, and when the job fails, the following retries fail because the copy script runs and blocks the backup file. Is there any possibility to implement it? If not, maybe it would be a new option for next release. What do you think?

Regards,

[v.benizano]

[merged]

Hello Everyone,

I've recently installed Veeam6 into our infrastructure, and was having some difficulty with a couple things. Any advice would be greatly appreciated.

- I have about 10 backup jobs setup (grouping like-VMs)
- Job # 1 is scheduled to run daily via the "schedule" configuration of the job.
- Job # 2 is run as a "post job activity" command, from Job #1
- Job #3 is run as a "post job activity" command from Job #2
- and so on.

1) Is there any way to configure the "automatic retry" on jobs that are run as post jobs? (I can configure automatic-rety on Job #1 as it is run as a schedule, this runs fine when there is a failed VM)

2) If I get a failed VM backup within a job that is run as a post-job activity, I can manually retry the job and it works great. But my only "problem" is that after the retry completes, the post-job activity of the job I'm retrying

will run regardless (which I don't want as they completed successfully the night before). So right now, for the job I want to retry, I'll disable the post-job activity, run the retry, after it completes, re-enable the post-job activity.

Are there any solutions/advice for these?

Thanks and Regards,

Vern

[robc.yk]

[merged]

Good Day folks,

We have a set of jobs that run in a chain. Each calling the next job through the post job run command. This all works fine but we have the following question.

How could we make one of the jobs run thier failed VM's again in the event there is a failure?

On a individual scheduled job the option is there to try X number of times after X mount of minutes, but how can this be done for each job that is run from a post run command?

Thanks in advance.

[dellock6]

Do you use this chain because you want no overlapping, or is there another reason?
Because if you only need no overlapping, is way easier to set cuncurrent jobs on proxies and repositories to allow only for one job at a time.

[robc.yk]

We have the chain done so that it is easier to schedule.
Essentially we have our jobs divided by departmental VM's and then further by SQL, Application, Oracle, or File servers.
Each job runs nightly with only 6 sobs with a scheduled start, the rest are chained together after the first 6.
On Friday's or Saturdays when the job is complete, the last one in each chain starts the HP DataProtector to get the files to tape.
This keeps only 6 jobs running across the 3 veeam servers at any given time.

So you are stating that I could set schedules for all the jobs and then use the concurrent setting to minimize how many jobs can run on each server? Would I have each job start at the same time, or have them a few minutes appart? How long is the que good for before something times out, or does it?

We have about 40 jobs now that run nightly.

*EDIT* I should also not that the last job that runs in a chain, when it kicks off the tape job, only backups the files included in the previous chain. Thsi way we can ensure that when the tape job runs, the backups for those files are not in use.

Thanks in advance for your insight.

[dellock6]

Hi Rob,
you can even set all backups to start at the same time, they will be executed one after the other if you set cuncurrent jobs to let's say 1.
To be sure the job with post-backup task invoking data protector, you can shedule it as the last one, giving them some delay, also because if you set all at the same time I'm not sure which order uses Veeam to run them.
I've never done what are you describing here, sounds fascinating but also a bit complex to manage, isn't it?

[robc.yk]

Interesting approach Dellock6,

I have changed all my jobs to start between 5:10 and 5:30 pm. Hopefully none will time out as they wait.
I am going to monitor the time it takes to finish them all over the next few nights. I will then think about how to best get it to tape.

I didnt think the way I was doing it was complex, but did require some documentation, perhaps that should have been my clue.

I have one question perhaps you or someone else can answer, We have a few proxy servers. They are using the EVA for thier storage. When selecting a backup, if the job runs on proxy 1 but is set to store its files on prox2 2 how is the data handled? Does a copy over the network happen? Is it best to tell the jobs to use the proxy that has the LUN storage it is assigned to?

Cheers all,

[dellock6]

If you leave all settings to automatic, proxy selection will be made by the Veeam server based on LUN access the proxies has,I always try to avoid to link a particular proxy to a job since in this way if that proxy dies the job does not run.
Also, what you are describing is a repository server, not a proxy; what you say happens because you have both proxy and repository roles on the same server: it's fine anyway, since proxy cuncurrent job does not count towards the repository activity, only check for server performances if it can handle both roles without slowing down.

Luca.

[jason252]

[merged]

I have some jobs with a post activity job setup (batch file / powershell file)

When a job fails, the post activity job is run, so in the batch file i check whether the job was ended successfully or not.
But when the job retries and finishes successfully the post activity job is not started again.

Is this normal behavior?

[jason252]

I see my topic is merged, but i have the exact opposite behaviour then the rest in this topic
The post activity job does not start when retry job finishes.

[Vitaliy S.]

Jason, then you need to open a support ticket, as I've just confirmed in the lab that the post backup job script is invoked for both "normal" and "retry" passes.

[muhaynes]

[merged]

Hi there,

I recently upgraded to v6, not sure if this was changed, but now if I manually cancel a job my post-job script runs. Also, if the job fails on a certain VM, the job knows to re-try 3 times, but it fires the script at the first failure. My post-job script sets off my tape backups, so this is a bit problematic. I've contacted Support and they've verified this is by design, but suggested I post here and see if we can work towards including an option to accommodate the re-try functionality with the post job scripts.

Thanks

[rhnb]

[merged]

Hi
I'm toying with the idea of using a post job activity command on our backup jobs to copy the backup to tape (using either TSM or Networker).

I was wondering what happens in the following circumstance...

Let say my job backs up 20 VMs.
The job runs, and 2 VMs fail to backup for some reason.
Does the Post job command run once this 'failed' job has finished in this case?

Now lets say I have it set to re-try the default 3 times, and lets say when the retry job runs it successfully backs up those 2 VMs.
Does the Post job command run then too?

Just trying to get my head around how I'm going to code this depending on the answer to the above.
I suspect the Post job command will run each time?

Also, I've got very little time to get this working, ( don't ask!), so any pointers to the best way to accomplish copying these backups to tape would be much appreciated.

Cheeers... Allan.

[lwhitworth]

[merged]

First a bit of background. I have three backup jobs that are chained one after the other, that is, each job calls the next as a post job activity. This is to ensure that the three jobs never overlap as the target is not very fast and two jobs running simultaneously would just upset it too much

Now, the issue I have is if for some reason a VM in the first of the jobs fails to backup correctly (but all others complete fine) and I have to retry the job, the retry backs up the one failed VM in the first job, but then calls the chained jobs as normal, so all VMs in the chained jobs go through a full backup cycle again.

What would be ideal, assuming it's not already possible and I'm just being thick, would be is if on the retry it either didn't run post job activities, or at least prompted to give the option of it running the post job activity or not.

Cheers,

Luke

[dellock6]

Set cuncurrency job limit on the repository to 1, so only one backup job will be executed at the same time, even retries.
After that, remove the chaining from backup jobs.

Luca.

[tmt]

[merged]

Hello,

We wish to run a PowerShell script using the advanced setting post job activity option. I know this is possible. However what happens if the backup job fails, does the command still get run? Also what happens if there are multiple servers in the backup job but you only want to command to run on a single server?

Many thanks.

[Vitaliy S.]

Also what happens if there are multiple servers in the backup job but you only want to command to run on a single server?

Post backup job script works for the entire job, not for VMs in the job.

[vnlbeheer]

Hi everybody!

We have 4 jobs in Veeam, each with multiple servers. The jobs are linked to each other, meaning that the first job is scheduled and the next one is linked to start after the first one finishes, and so on. However last night a backup of 1 VM failed in the first scheduled job. Running a retry resulted in the failing VM being backupped, but also in kicking off incremental backups of the linked jobs, which were all completed succesfully. If I do a retry of a job that has failed, I would expect it to just do that job, since it's not a regular scheduled action being kicked off.

Is this known/wanted behaviour or is it something that was overlooked and will be fixed? Running Veeam B&R 7 Enterprise with patch 1.

Thanks!

Jim

[foggy]

Jim, what is the reason of chaining jobs? If you are limiting the concurrent load on your proxy/repository servers this way, then the recommended approach is using max concurrent tasks numbers instead. In this case there is no need to chain jobs, and there will no issues with starting jobs manually.

See these topics for further reading > Quick Question - Concurrent Tasks and Veeam B&R 6.5 jobs chain.