Veeam warns adding Backup Repository - HP StoreOnce Catalyst

[wanderer]

Hello,

Just wondering if this Veeam message can be safely ignored?

Number of devices per initiator port in Catalyst over FC settings is too low.
We highly recommend you to increase this number in StoreOnce management UI
Proceed anyway?

Because we tried increasing the number in StoreOnce and this messages still shows while adding the StoreOnce Catalyst backup repository.

Thanks.

[Mike Resseler]

I am not sure if this can be safely ignored, it will have certainly some impact on performance

The number of StoreOnce Catalyst over Fibre Channel devices available must be at least one more than the number of streams if the backup application or plug-in can open command sessions during a backup or restore.

There is more information on this in this guide: https://www.hpe.com/h20195/v2/GetPDF.as ... 336ENW.pdf

Hope it helps

Mike

[CraigL2112]

This error cannot be ignored. If you try to run more streams than the # of devices per initiator port are set at, Veeam will throw a -1404 OSCLT_SERVER_OFFLINE message and your backups/BCJs will fail.

Increase your # of devices in your StoreOnce console, then on your Veeam server, refresh/rescan the hardware in Device Manager. If you expand your 'Other Devices' section, you'll see many "HP StoreOnce CoFC SCSI Processor Device" entries. The number will be equal to the 'Number of Devices per Login' setting multiplied by the number of ports you have zoned into your Veeam server. For example, if you have 16 as your 'Number of Devices per Login' value and you have both StoreOnce FC ports zoned in, you'll see 32 entries for 'HP StoreOnce CoFC SCSI Processor Device' entries.

We just hit the error I stated above recently when we upgraded our StoreOnce code to 3.15.1. That 'Number of Devices per Login' value was reset to 1, which caused our jobs to begin failing immediately. HP knows this is an issue, and the fix is simply what I am suggesting you do

Let me know if you have any questions at all and I can try to help.

-Craig

[wanderer]

Mike, thank you for the direction.
Craig, thanks a lot for the detailed, clear and friendly explanation!

The setting "Devices per Initiator Port" in our StoreOnce is 4 per FC port (as shown in the below picture). That would double to 8 as you explained because we have two FC ports zoned into our Veeam server, and we should see 8 "HP StoreOnce CoFC SCSI Processor Device" entries in Device Manager. (Please correct me if I am wrong.)



In light of the above, our Veeam Catalyst Backup Repository setting "Limit maximum concurrent tasks to" should be set to 8 instead of 4 (our current setting). If that number is set too low, Veeam would warn us "number of devices per initiator port in Catalyst over FC settings is too low" when we proceed to add the backup repository. (Correct me if I am wrong.)



Strangely, unless my memory fails me, I remember we did increase this number in Veeam and the message still prompted though.

I will revisit the infrastructure later next week, so I will try again and report by then.

Thanks!

[CraigL2112]

You are spot-on with your assessment. Give it a shot and report back. FWIW, I have mine set to 32 per port.

The good news is that you will know you are hitting the StoreOnce stream cap if you get the -1404 error I referenced previously. Additionally, I'm pretty sure that just the fact that you got that warning means your Veeam -> StoreOnce security is also set up properly.

On a side note -- if you have a moment, give the StoreOnce 3.15.1 release notes a look. In addition to a total GUI overhaul and a slew of firmware upgrades (at least for our 4500s), we are unquestionably seeing higher throughput w/ CoFC than with 3.13.4.

-Craig

[foggy]

In light of the above, our Veeam Catalyst Backup Repository setting "Limit maximum concurrent tasks to" should be set to 8 instead of 4 (our current setting). If that number is set too low, Veeam would warn us "number of devices per initiator port in Catalyst over FC settings is too low" when we proceed to add the backup repository. (Correct me if I am wrong.)

Veeam B&R actually warns if the number of devices on the gateway server is less than 64, regardless of the repository tasks limit. So you need to increase the StoreOnce "Devices per Initiator Port" setting to at least 32 to avoid warnings.

[wanderer]

Hello again with much thanks to Craig and foggy (Alexander) for the clarification!

Our Veeam server has 16 CPU cores (and 96GB RAM), according to https://helpcenter.veeam.com/backup/hyp ... tasks.html :

To use this capability (edit: Parallel VM processing), proxy server(s) should meet system requirements – each task requires a single CPU core, so for two tasks to be processed in parallel, 2 cores is the recommended minimum.

If that is true, it means we better configure Veeam to "Limit maximum concurrent tasks to" 16 (same as our no. of CPU cores) to meet this recommendation.

At the same time, we should configure StoreOnce "Devices per Initiator Port" to at least the same number, i.e. 16 (actually 8 because of 2 FC ports) for us, to avoid the -1404 error Craig encountered.

According to foggy (Alexander), with Veeam "Limit maximum concurrent tasks" configured as above, it means we cannot avoid the Veeam warning. However, it should not be an issue.

Please correct me if I am wrong. Thanks again!

[foggy]

You're correct in your considerations. If you set the number of devices to 32, you will avoid warning as well.