SOBR Tiering Job Fails with "Shared memory connection has been forcibly closed by peer"

[marius402]

I am experiencing this error on some servers doing SOBR tiering:

"Processing XYZ Error: Shared memory connection has been forcibly closed by peer."

It happens to multiple jobs in the same SOBR tiering level, but also across different sites. Some sites works without any issues and some sites fail with this and does not upload anything to Azure Blob.

The case # 05106091

Regards

Marius

[veremin]

The support assumption is that application is crashing due to resource overload, there are some ways to address it, including increasing server memory, decreasing repository concurrent task limit as well as number of HTTP requests processed within each task.

Kindly, apply the steps proposed by the support engineer and see whether it makes any difference.

Thanks!

[marius402]

Thank you Vladimir, I have received the post and will work through it today and feedback

[DeanCTS]

I was also doing some work on this one today based on Vladimir's recommendations about decreasing concurrent tasks.

[veremin]

Are you getting the same error, or are you affected by a different issue discussed in the adjacent thread (slow performance of object removal process spotted in several Wasabi regions)? If the latter, it's recommended to work with the Wasabi support team directly. Thanks!

[TonioRoffo]

The good thing about object storage though, is that the upload resumes in the next SOBR offload task and finishes without problems.