Thoughts about Veeam Ready - QuObjects

[kristofpoppe]

Hi all,
Wondering how your experience is with QNAP's QuObjects implementation.

We have several devices, fully Veeam Ready for Object with QuObjects.
Small backups work without any problem, but as soon as we start with lager chunks of data, communication suddenly stops. No data is sended to the repo, no time-outs are reached. This ends with jobs running for more than 60 hrs without any progress. They'll run forever.

Curious sidenote:
When we open the properties of the repo via Backup Infrastructure > Backup Repositories and then just click through the settings (Name, Account, Bucket, Mount server,...) The job starts sending data when we go from Account to Bucket (at that time, there's communication with the object implementation on QNAP's side).
It seems that this action 'revives' the repo for a short period of time.

We opened a support ticket (Case # 07262669 ) but apart of entering a regkey "TcpMaxDataRetransmissions" (which didn't help) we're out of solutions...

This is a straight forward implementation of local QuObjects S3 compatible storage on-prem.
I've read several other posts that people are facing a lot of problems with the QuObjects. These devices should then lose their Veeam Ready status.
"It just works!" applies for the backup software, but not for the alligned storage...

[david.domask]

Hi Kristof,

Thank you for sharing the case number and your observations, and sorry to hear about the challenges. I can see that the case was opened only recently and logs just got uploaded, so while it's unfortunate adjusting the MaxTCPTransmissions value did not assist, I think let's allow Support a bit of time to review the logs and further develop the action plan.

The observation about the throughput increasing a bit when you edit the properties is a bit unusual; can I ask you to please include this information on the case also if it's reproducible? (Noting the date/time of the tests would be useful to help match it to events in the logs, just a few quick demonstrations, but keep in mind this might not be related to Veeam specifically, so the logs may not reveal much, but it sounds like a pretty fast test and it would help to narrow the focus a bit).

Regarding NAS devices and their S3-enabling Apps, they can work in a stable fashion but there are of course always some caveats. Just for clarity purposes, are your QNAPs specc'd similarly to the one from the Veeam Ready Page? With SSDs and similar or better hardware? Since you mention that the jobs run "okay" and then hang at 99%, I do wonder if it's related to retention; this can be a fairly involved operation on S3 and the deletes may be what are taking a bit of time, but Support will be able to shed some light on which operation is happening when it's at 99%.

Let's see what Support can find here and the action plan that develops after the log review.

[kristofpoppe]

Hi David,

Thanks for your quick reply ! I'll check if I can reproduce the behaviour when 'touching' the storage.
I'll wait for the feedback from support team. On the other hand, I can confirm the boxes meet the hardware requirements.
Corei5 CPU, descent amount of memory, full SSD.