Frequent 501 errors when running MPS

[Kristen-Z]

Hi, we frequently encounter HTTP errors when running the MPS service on VRS files. Sometimes all files fail with 501, sometimes only one VRS fails while the others succeed. We’ve attached screenshots and logs from two failed runs.

2025-12-17-15-49-41.log

2025-12-17-15-37-50.log

Do you know what usually causes this, and how we can fix or avoid it?
Let me know if you need more information. Thanks!

[ja-martinez]

Hello @Kristen-Z, thank you for bringing this to our attention. I'm going to research some things and get back to you.

[Tarunstk]

Hi @Kristen-Z the error 501 usually means the MPS has failed due to health check failures. I see you have trimmed the sequences and are trying to process them. How have you trimmed the sequences? Are there any sequences that you have trimmed inconsistently?

And could you tell me why you are trimming the sequences?

[Kristen-Z]

Hi, thanks for the reply!

We’re trimming the sequences using the official tool, e.g.
vrs copy original.vrs --to new.vrs --after +2

The 501 errors seem fairly random. The sessions we’re processing have all passed the health check, and we haven’t noticed any inconsistent trimming across sequences. In our experience, failed health checks usually lead to error 102 instead.
Are there other possible reasons that could trigger a 501 error, and is there any recommended way to fix or avoid it? Please let me know if you need more information.

[ja-martinez]

@Kristen-Z I see, let me do some more research and get back to you. Thank you for clarifying.

[Tarunstk]

Hi @Kristen-Z could you request MPS again on the same files. I wanted to check in and see how you trimmed those vrs files? Was it from the beginning of the vrs files? It seems like you are trimming them from the beginning.

As for Error 102 goes you will be able to find more information on that from the health_check.json files and the summary.json for each run. Please let us know if re running it succeeds?

[Kristen-Z]

Hi, thanks for the reply!

I re-ran MPS on the same files and added the --retry-failed option. This fixed most of the 501 errors, but one of the recordings is still failing with error 501 for SLAM MPS.

Regarding trimming, yes, we’re trimming from the beginning of the VRS files using the official tool (--after +2). Could you suggest any additional steps to resolve the remaining error, or any specific checks we should run?

Thanks for your help!

[Tarunstk]

Hey @Kristen-Z okay the positive thing is that most of the recordings were processed after re-running with the --retry-failed flag.

As for the one file that is failing the SLAM sequence let me look into that and get some more insight into it get back to you. Thanks for your patience.