-
Notifications
You must be signed in to change notification settings - Fork 51
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
WIP: Retries #232
base: main
Are you sure you want to change the base?
WIP: Retries #232
Conversation
Still need to do sleep/backoff, etc. Hoping to remove NodataReader too.
When an `open` fails and we'd return nodata instead, we don't cache that anymore. So if the asset doesn't exist, we'll try to open it for every chunk. Unclear how much of a performance impact this will have. It's probably not ideal, because combining it with retries could be nice: retry errors first, then give up and use nodata if they persist?
got lost in #221, which was a reasonable change, but turns out there was still a test importing it
TODO: - sleep & backoff - figure out how to handle sleep in tests
Hi @gjoseph92 thanks for this feature. I had a look an although I have not tested the code it looks good. The way of using this feature seems versatile so imho, nice, I would defiantly use this feature! I see that you have not included the Also, I would think that the possibility to add a delay between retries would be also useful. |
Hello @gjoseph92 . Do you think you will merge this anytime soon? |
Support retrying errors matching a given pattern.
This also moves error handling logic (both retires and nodata) up to the
to_dask
level, and out of Readers.Closes #18.