Open vsnine opened 5 years ago
This was seen against DO container 1.5.0.
Agreed. We generally have to retry on errors in DO because the VE may still be coming up. However, if we get a 401 we can stop trying (or at least shorten the retries from there)
AT-478 for tracking
Found that typoing the password of the BIG-IP, which had been changed previously during bootstrap, the declaration takes a long time to abort. Digging into the container logs for restnoded, if bigIpPassword is incorrect, the 401 error is seen but the declaration doesn't immediately fail. The system continues to retry authenticating to the BIG-IP. Making subsequent requests for the job status just continues to display code 202 with message "processing".
Restarting the container is a quick fix to stop the retries.
Shortly after the initial failure it indicates that there are 90 attempts remaining and continues to try with the bad password. Example:
If I allow it to continue until it runs out of retries, this is the result.
At this point requesting the task status does return a 500 code along with the error message seen initially.
I feel like this should abort after one or two retries if not immediately.