High Latency

[oryx360]

Hi,

We are seeing high latency on our Cisco UCS system (VSphere 6.5). Our backup person runs the Veeam Backup and Replication and we have been notifying high latency for the last couple of months. I want to check one by one to see which is causing the issue. Whether it is the datastore side, Veeam backups and replication or Cisco UCS side or the Network side. So, I would like to get some detailed information where I could pinpoint the problem and look in Veeam to see whether I can find any information or whether I can look into the VSphere side and point it out.

Our Veeam backup and replications goes through the day sometime due to unfinished backups before 6am.

How can we check where the problem lyes.

[Gostev]

Hi.

You should really work with Cisco on this, they are in much better position to tell you what component of their infrastructure is struggling with the load.

While Veeam provides real-time bottleneck analysis (see sticky FAQ), it is designed to find bottlenecks in backup infrastructure components and is not granular enough for your need. For example, since you have high storage latency, Veeam will likely tell you that the bottleneck is "Source" (datastore) because reads take too long to complete and most of the backup job time is spent waiting for the requested data. But we can't possibly know if its actual storage or storage fabric that is causing these delays.

Thanks!

[oryx360]

Hi Gostev,

I did go through some reports through Veeam. I see one report that really caught my attention. The Orphaned VM snapshots. So, we do not se any VM Snapshots in VSphere, but do you think that these still reside in the datastore?

Thanks in advance

[Gostev]

Sure. This is exactly what makes them "orphaned"