CDP Alerting

[nhefling]

I have a customer that is using CDP and the alerts are impossible to understand. for example. Alert email subject line is Failed....top line of CDP job status is Warning and the last session status is success. How is anyone suppose to read this and understand if they should be worried? What is Veeams next release going to contain for alerting information?

[Gostev]

Please share a screenshot of the email so we can better understand the specific situation.

Overall it does not seem too different from how backup job email reports worked since inception:
1. Entire job will be marked as failed if processing for at least one machine has failed. This situation requires an immediate action from backup admin, because no restore point was created. They should be worried.
2. Email report may in addition contain warnings for less critical issue (for example, backup repository is running out of disk space). These require an action too so they should be worried here too.
3. Most machines can be at the same time market with success in the same report, because they were backed up successfully.

In other words, they should be worried unless their report is all green. This requires working through all errors and warnings to eliminate their root cause. Some warnings such as for RPO violation can be simply disabled in the CDP policy settings in case they don't care to be notified about these types of issues.

[nhefling]

I cannot find a way to easily paste a screenshot here. Can I send directly to you?

[Gostev]

Just share it on any screenshot or file sharing platform, and post a link to the image.

[RubinCompServ]

Sometimes Veeam notification emails say Failed even if the most recent session completed in Success, if there was a previous session that contained a Failure. I opened a ticket with Veeam for this and they said it was by design.

[veremin]

Do you mind sharing the ticket number so we can check the issue and its resolution? Thanks!

[RubinCompServ]

Case #05241960

[veremin]

We have clarified the situation internally and what you see is indeed the expected behavior:

CDP activity consists of different sessions (from the smallest to the biggest entity):

• VM sync session (time period defined as RPO in CDP policy settings)
• VM period session (time period defined as long-term interval in CDP policy settings; consists of number of VM sync sessions)
• Policy session (time period is 24 hours; consists of number of VM period sessions)

If VM sync session fails, its parent VM period session gets failed status and so does policy session.

A policy session that got failed status during 24 hours at least once is considered to be a failed one, even if the latest VM sync and period sessions are successful - and this is exactly what your daily report illustrates.

Thanks!

[RubinCompServ]

@veremin

I assume that was a response to the orginal question (@nhefling), as my case wasn't related to CDP notifications, but backup notifications.

[veremin]

Yep, for some reason I've assumed you were also using CDP.

For backup job it's unexpected to receive the failed job state report, if the last retry was successful and suppress notification until the last retry option was enabled.

If you experience this issue, kindly re-open existing or create new ticket and we will try to see what can be done from our side.

Thanks!

[RubinCompServ]

It just happened again last night, as a matter of fact.



I had a ticket and it was closed, saying that this was by design, so that customers would be aware that there had been a failure and could review/repair whatever the transient issue was. The fact that it makes it more confusing to people to see Failure alerts in the morning but no actual failed jobs didn't seem to factor into the design decision.

[veremin]

I cannot open the image, unfortunately, but can you tell me that both conditions were met: the last retry was successful and suppress notification until the last retry option was enabled?

If so, kindly collect debug logs and open a new ticket - we will ask R&D team to review the case, cause the situation does not look expected.

Thanks!

[RubinCompServ]

I can confirm that the last retry has completely successful and the option to suppress notifications until the last retry is enabled. I'll get that ticket opened. Thanks!

[veremin]

Got it, and the image-hosting server is online again, so we can see the screenshot. Based on the image provided, it looks strange that VM that had successful retry session is still identified as failed one.

Kindly, provide a new ticket id and short issue description within new thread, it will help us not to derail the existing discussion any further.

Thanks again for understanding, and we are looking forward to hearing from you.