"Waiting for resource availability" troubleshooting

[lrosa]

Veeam jobs are often waiting for resource release, this is normal.

In a trivial setup is easy to tell which resource is blocking what.
In a complex setup with jobs, copy jobs, proxies and so on sometimes it requires a lot of clicks to locate the job that is blocking the others.

Is there a way to query some Veeam structure/table in order to have a list of who is using what?
For instance, given a proxy with 4 concurrent jobs, the query command should return a list of 4 rows with the Job/VM/vDSIK

Think of it as a sort of task manager of Veeam.

Consider that my scope is realtime information, not statistical or report oriented.


Thanks!


Luigi

[PetrM]

Hi Luigi,

This information can be found in Veeam debug logs.
However, if you notice that jobs are waiting for resources for a long time, you may contact our support team, they will help you with troubleshooting.

The basic requirement is to adjust proxy and repository slots according to the number of tasks you're going to run in parallel.

Thanks!

[lrosa]

Thank you!

A sort of "task manager" real time could be the tool to better size the proxies and repo slots.

Luigi

[PetrM]

You may take a look at RTS.ResourcesUsage.log (C:\ProgramData\Veeam\Backup), it logs resource usage in real time.

Thanks!

[markhensler]

You may take a look at RTS.ResourcesUsage.log (C:\ProgramData\Veeam\Backup), it logs resource usage in real time.

Thanks!

I just learned about this on a support call. Very cool.

I also would be interested in seeing something added to the Veeam Console UI that indicated usage/capacity per Proxy and Repository/Extent.

I would be even more impressed if there were a real-time/last-24-hour report listing the running tasks (as opposed to jobs, which we have) which included proxy and repo/extent as sortable/filterable output columns.

And while I'm dreaming... It would also be helpful to see a report that showed the average wait time for each task reported by proxy and repo/extent.

[markhensler]

I currently have 33 tasks running, most of which are waiting for resources. Yet, my RTS.ResourceUsage.log file looks like this:

Code: Select all

[30.04.2020 16:49:50] <381> Info     |--------------------------------------------------------------------------------------------------|
[30.04.2020 16:49:50] <381> Info     | Resource name                  | Usage | Id                                                      |
[30.04.2020 16:49:50] <381> Info     |-------------------------------------------------------------------------------- vSphere Proxies -|
[30.04.2020 16:49:50] <381> Info     | jzveeam-bu2                    | 1     | {49868557-2ed4-4234-a0ce-451d56ab4616}                  |
[30.04.2020 16:49:50] <381> Info     | jzveeam-bu5                    | 1     | {3b7f83eb-6e95-488d-8d35-e9358975c6c6}                  |
[30.04.2020 16:49:50] <381> Info     |----------------------------------------------------------------------------------- Repositories -|
[30.04.2020 16:49:50] <381> Info     | JZVeeam-BU5 Local-5a           | 2     | {9bb69b4c-ec92-49ba-b279-8239a299163d}                  |
[30.04.2020 16:49:50] <381> Info     | JZVeeam-BU6 Local-6a           | 9     | {b979eaec-2ac0-40ca-83b8-299205e0858f}                  |
[30.04.2020 16:49:50] <381> Info     | JZVeeam-BU4 Local-4a           | 5     | {d537a7c0-2428-4e66-a31b-9ac79515644d}                  |
[30.04.2020 16:49:50] <381> Info     | JZVeeam-BU3 Local-3a           | 5     | {3c0225e4-7de5-4138-b272-6a86f13d14b9}                  |
[30.04.2020 16:49:50] <381> Info     |--------------------------------------------------------------------------------------------------|

Why doesn't the sum of the Proxies match the sum of the Repositories?
What tasks occupy a task slot for Repositories that does not also occupy a task slot for Proxies?

[Gostev]

Proxy is only required for tasks which interact with VMs (either pull data out of VMs, or put data into VM disks). There are plenty of features that don't interact with VMs at all, only with backups - and thus only need repository slots. For example, to name a few: Instant VM Recovery, Backup Copy jobs, Capacity Tier offload.