hail
hail copied to clipboard
[batch] Open Batches!
Based off of #12154. Stacked on #12193.
@daniel-goldstein Let's see if the tests pass now, but should be really close!!!
I forgot to add the UI time_updated field... Will do so now.
I put the time_updated field in. I'll test it with dev deploy once everything else is ready to go.
Stacked on #12208
Note to self to check the logs for error messages before merging!
I changed that the _update_token
is no longer cached. I thought it was the source of a bug, but it was something else that was an issue. However, I felt it was confusing and I didn't see what value it provided as we have retries on all of our client operations that would need the token and the token is not used in future operations to submit the update.
duh you're totally right, I don't know why I thought retrying was happening higher up and not further down in the http client.
Sorry this was a lot more broken than I thought. I didn't remember everything I stripped down for the previous PRs and didn't add back in. The commit update and update-fast endpoints need to return the start_job_id
. Hopefully that's the last of these issues