Custom Alarm for Backup/Replication job duration

[chall]

Hi,

I'm trying to setup a custom alarm to alert me whenever any of my Veeam B&R backup/replication jobs take too long to run. This will give me an indication of any slow throughput for example and not just when the jobs are failed/successful.

I've configured a new rule using "Job duration exceeded the allowed timeout period" and configured a custom duration but then I am prompted to choose a VM.

As I have a lot of VM's it would be easier if it was possible to do this for all VM's with a wildcard or something? This would also cover any new backup/replication jobs I create in the future.

Additionally - are there any other ways I am missing that would help me to monitor the performance/throughput of the jobs I'm running?

Appreciate any advice you might be able to give.

Cheers
Chris

[Vitaliy S.]

Hi Chris,

Actually, alarms creation wizard prompts you to select the job name and not the VMs added to the backup job.

As regards job performance/throughput data, then you can review the latest state (duration and transferred data) of the backup/replication job via Jobs tab on the corresponding backup server. If this information is not exactly what you're looking for, then please elaborate your question further. Any thoughts on how to enhance our current functionality are highly appreciated.

Thanks!

[Vitaly.Mzokov]

Hi Chris,

I've configured a new rule using "Job duration exceeded the allowed timeout period" and configured a custom duration but then I am prompted to choose a VM.

As I have a lot of VM's it would be easier if it was possible to do this for all VM's with a wildcard or something? This would also cover any new backup/replication jobs I create in the future.

I am sure you are saying that the rule has asked you to specify a JOB name. You may feel free using * symbol as a wildcard for a "Job name" parameter.
Using this wildcard you will trigger an alarm for any type of Veeam Backup activity based on job duration criteria you have defined.

Hope it helps!
-VM.

[chall]

Thanks Vitaliy, You are correct. I did mean JOB name not VM - I have tested using * symbol and that works for any job.

But I see it only triggers the alarm once for the first job - so if subsequent jobs run past the specified time I won't receive another alert until I have manually resolved the first one.
The only way I can see to work around this is to set up individual alerts for each job, but that will have to be managed as I add future jobs.

Unless I can use suppression options to achieve what I need?

Thanks again for your support.

Chris

[Vitaly.Mzokov]

Chris,

Yes, you're right. I was able to reproduce that behavior in my lab. I will pass that information to our R&D team for further research. Thanks for pointing that out.

-VM.

[Vitaliy S.]

Chris,

But I see it only triggers the alarm once for the first job - so if subsequent jobs run past the specified time I won't receive another alert until I have manually resolved the first one.

No exactly, you should receive email notification on the breached threshold for other jobs too, however only one combined alarm will be visible in the backup console, it is by design. If you need to review all jobs that has exceeded their max duration, please double-click on the alarms history to review all jobs. Anyway, I will verify the behavior you describe with our QC team.

The major use case for this alarm is to track individual job duration. I believe you would agree that jobs duration can vary a lot depending on the number of VMs being processed, thus it's hard to set a single threshold for all jobs, that is why you should be using separate alarms for different jobs.

BTW - how many jobs you have in your environment?

Thanks!

[chall]

Vitaliy,

No worries - I look forward to any solution you can offer. The new Data Protection View functionality you've added into Monitor for v6.5 is really useful (and logical!), I am keen to make the most of it.
Can you keep me updated with any progress the R&D team make on this? I will of course lookout for any updates/patches that get released.

Thanks again.

Chris

[Vitaliy S.]

Chris, you need to open a support case to be notified on all patches released by our R&D team. Also be aware that Anton usually informs all registered forum users on our product updates in his weekly digest email.

[Vitaly.Mzokov]

Right! You may use our Customer Portal to open a case. This is also a nice chance to ask other questions regarding the product and get an immediate assistance!

-VM.

[chall]

Chris,
No exactly, you should receive email notification on the breached threshold for other jobs too, however only one combined alarm will be visible in the backup console, it is by design. If you need to review all jobs that has exceeded their max duration, please double-click on the alarms history to review all jobs. Anyway, I will verify the behavior you describe with our QC team.

The major use case for this alarm is to track individual job duration. I believe you would agree that jobs duration can vary a lot depending on the number of VMs being processed, thus it's hard to set a single threshold for all jobs, that is why you should be using separate alarms for different jobs.

BTW - how many jobs you have in your environment?

Thanks!

Vitaliy,

I have seen the history of the job shows which VM's are triggering the alarm. But as before once the first job triggers the alarm no further jobs that exceed the specified threshold can trigger that alarm until you resolve it - therefore the history does not show which jobs have breached AFTER the first alarm is triggered.

I have 34 jobs configured in house (mostly indivual VM's, but some jobs with multiple). Generally none of the jobs run for much more than an hour in my case so I wanted to set a global threshold for this alarm of approx 2 - 4 hours so that I will be made aware of any issues that would impact on the performance of any of the backup/replication jobs.

If the jobs suddenly start taking more than 2 hours, I know I have an issue somewhere!

This week we had a problem with our core switches which had a massive impact on Veeam's performance. I was seeing very low throughput (approx. 100Kb/s instead of 100+Mb/s) but because the jobs weren't failing I wasn't aware for several days until I logged on to the Veeam B&R server by chance!
I am looking to be able to monitor this performance in house but also for some of our customers who already have Veeam B&R but aren't using Veeam One. Some of the customers who I will be monitoring this for will have many more backup jobs than I am currently testing with here.

Hopefully that will help you understand where I am coming from

Thanks
Chris

[chall]

Thanks I will raise a support call if neccessary, I just thought I was missing something easy!

[Vitaliy S.]

This week we had a problem with our core switches which had a massive impact on Veeam's performance. I was seeing very low throughput (approx. 100Kb/s instead of 100+Mb/s) but because the jobs weren't failing I wasn't aware for several days until I logged on to the Veeam B&R server by chance!

But wouldn't job duration alarm detect this kind of issues? If, all of a sudden, your jobs start to crawl, this will immediately be spotted by the job duration alarm, and you will be notified about poor job performance anyway, right?

who already have Veeam B&R but aren't using Veeam One.

In this case they will need to start using Veeam ONE