Add API retry logic to all Sindri API Calls #118
Assignees
Comments
sangaline
added a commit
that referenced
this issue
Jul 24, 2024
This adds a second argument to the `SindriClient()` constructor which can include a `retryOptions` field which controls the newly added retry behavior. The options are passed through to the `retry` package and are documented [here](https://www.npmjs.com/package/retry#retrytimeoutsoptions). By default, we'll retry 4 times with exponential backoff starting at a 1000 ms timeout. The logic to determine whether to retry is hard-coded, we retry for status codes of 502-504 or any sort of connection errors. Note that we use a fork of `p-retry` here which supports both ESM and CJS because we provide builds for each in our package (see: sindresorhus/p-retry/issues/76). Closes #118 Merges #120 LGTM given by: @katiemckeon
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add API retry logic to all Sindri API Calls
To mitigate common api connectivity issues, we should implement simple retry logic for 502, 503, and 504 error codes.
The sindri python SDK already implements this retry logic. Please use it as a reference.
https://github.com/Sindri-Labs/sindri-python/blob/5da072e1631a13532fbacabfc550ba2b11fe1669/src/sindri/sindri.py#L239-L248
The text was updated successfully, but these errors were encountered: