DiamondLightSource / DiamondLightSource/mx-bluesky
Hyperion should be more tolerant of transient, retriable HTTP errors when calling services
Open
enhancement
hyperion
- Dominant language
- Python
- Stars
- 4
- Forks
- 5
- Avg merge
- 1d 30m
- Merged PRs (30d)
- 2
Description
Hyperion should be more tolerant of errors that it can retry, such as this incident calling agamemnon:
https://diamondlightsource.slack.com/archives/C017UDM8R5W/p1772696004966989
We should go through and look at our calls to expeye, agamemnon etc and try and make them more robust rather than causing collection to fail straight away.
## Acceptance Criteria
- When errors are encountered we may attempt a limited number of retries, before ending UDC and returning the baton if the retries fail.
Contributor guide
Assessment
This issue has not been assessed yet.