Storage Integration, DirectSAN with Linux Repository and Netapp over FC

[DaStivi]

Hello there, anybody having ah good advice for me, getting following error trying to rescan my Netapp Appliances:

Error Failed to obtain list of VMs on LUN vmstore.lun of volume fcp1_vmstore - Error: Cannot find ONTAP initiator group for any existing backup proxy server

so volumes and vms get gathered correctly from the netapp "admin" (web) interfaces, but as soon as the volume clone/map/rescan operations should start it fails..

using an RHEL Linux HPE Apollo machine with FC Adapters, SAN zoning etc. does work, as i can manuall map volumes to the Linux Servers, i can rescan the hba's i'll see the disks in fdisk and also multipath tool... i even think its not even that far that the linux proxy trys to connect to the netapp volumes it fails allready before...

i didn't find any more requirements regarding "software" packages that maybe veeam needs in rhel/linux to be able to gater hba informations, rhel is installed as minimalistcly as possible, only added the dm-multipath and sysfsutils

in storage rescan veeam looks like to scan what igroups are configured on the netapp and "compares" ? it to information in its database? so either the netapp does not provide the information to the rescan process to veeam, or in the database of veeam the ww(p)n's of my hba's are missing? idk...

in logs there isn't any more helpful hint other then the same line of error above...

[HannesK]

Hello,
That sounds like a technical issue. Please provide a support case ID for this issue, as requested when you click New Topic.

Without case number, the topic will eventually be
deleted by moderators.

Best regards,
Hannes

PS: support can only help if you upload logs https://www.veeam.com/kb1832

[DaStivi]

opened case and upload logs right now:
Case # 05027931

[DaStivi]

no one using Linux Repos with Storage Integration? still trying to find ah solution with Support, but not got any step further...

[HannesK]

Hello,
sorry for the inconvenience. I just checked the case and it looks like the engineer did not realize that it seems to be a known issue which happens in specific scenarios. So it's not a general issue and chances are low that someone external can help with that.

The case should be escalated soon.

Best regards,
Hannes

[DaStivi]

i think i found the issue today! gonna check everything now and will also inform support...
basically it's not directly ah bug, but VBR could also made it better to not stop the host rescan job instead ignores it...

somehow on the HPE (Apollo) Server there is ah SD Slot, and Linux does recognize this slot and showed it as /dev/sdc, host rescan in veeam failed with "no medium" -> same message as when trying to mount device manually... sure no sdcard was in it, it did costs me some time to find out the /dev/sdc was the sd card slot, after disabling the sdcard slot in bios of the server the sdc went free and error in veeam also disappeared and rescan of host and subsequent storage/vol-lun rescans also started to working!

only thing that i can just speculate, i've got 2 identical hosts, only one had the issue for now, but also this was the only one with the proxy role allready on it.. so it might be just the case that the host rescan, rescans volumes/disks only when proxy service is installed...

so either way, disabling sdslot helped but in my opinion also the host rescan should ignore the failure instead of aborting the rescan job!

for the search engine, this was the error messages i got in the host discovery job:

[servername] Refresh Linux host failed Error: No medium found
POSIX: Failed to open file [/dev/sdc].
Agent failed to process method {DiskManagement.QueryScsiDiskPath}.

[servername] Disks discovery failed Error: No medium found
POSIX: Failed to open file [/dev/sdc].
Agent failed to process method {DiskManagement.QueryScsiDiskPath}.

[rennerstefan]

Hi,

glad you found the answer and thanks for posting the solution here.
We will also check in our labs.

Thanks again

[Ollo]

Hello, we have the same error here.
Where did you disabled the sdslot? In ILO or in linux?

[rennerstefan]

after disabling the sdcard slot in bios of the server the sdc went free and error in veeam also disappeared

Seems the best way is to disable it in BIOS.
With that the OS will not see it at all.

[DaStivi]

Hello, we have the same error here.
Where did you disabled the sdslot? In ILO or in linux?

yes disabled SD Slot in BIOS! i think this should match, its kinda hidden in the menus..: https://techlibrary.hpe.com/docs/iss/pr ... _slot.html
its the "internalSD" Slot thats located on the Systemboard, where someone might install ESXi on that SDCard (what you shouldn't do either ) but basically its enabled and linux does display this device... didn't noticed it in Windows, or at least windows doesn't display the "empty" Device with ah Driveletter...

[Ollo]

Our server was an windows server before. windows doesn't has the problem with rescanning. We try to disable it in bios.
@Veeam, changing sourcecode is needed, that rescanning is possible with an active sdslot and some informations in Documentation Center might be helpful.

Thanks

[HannesK]

yes, it's tracked as bug #291165 as far as I see, but it's not fixed yet.