My save to repository task is throwing this error. How do we troubleshoot and resolve this?
We are currently experiencing some instability in business processes. Please open a support case or check the status page for updates.
This is not related to the object reference error which started just the other day, it started on 8/31/2022 and has not been resolved since.
Are you not able to resume it?
No, I found the problem today while looking at other errors, see it has been a problem since last month. I tried retrying and it is still failing with the same error. I also assume it retries on it's own interval.
Please open a support case and provide the customer ID and the ID of an affected instance.
They went through after the server maintenance, but it seems Cloud is not retrying on intervals which can cause these archives to sit for months if they happen to be run while a service is having problems.
Right, but we'd still want to track down what happened to cause to get stuck in the first place.
Support Ticket: 229532
Support checked on the backend and found that between the Business Process service and the Repository service there is a SAML based authentication configured. At the time the Repository system could not provide a SAML token for an unknown reason. This prevented the BP service from sending data to the Repo service.
The concern here is that the files were never archived due to this issue and it is not the only occurrence.
I asked why, when I finally discovered that it was suspended and tried to retry by pressing the Play button, that it failed with the same exact error at that time as well. They said it was coincidental that the repository service was down for other reasons at this time (I was in there troubleshooting other repository and workflow issues when I found these). However it is confusing as to why it threw the same exact error if the problem was coincidentally something else.
The other concern is that the only way I found these was with a temporary access account, SupportUser, to troubleshoot another issue. Was the customer meant to be notified of something like this? Or even better, if there is a SAML issue between internal services can the system re-process these once the token is available and not remain stuck for months on end?