Add RPC Retries via interceptor #24
Merged
+260
−18
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
cadence/_internal/rpc/retry.py
Outdated
|
|
||
| def is_retryable(err: CadenceError, call_details: ClientCallDetails) -> bool: | ||
| # Handle requests to the passive side, matching the Go and Java Clients | ||
| if call_details.method == b'/uber.cadence.api.v1.WorkflowAPI/GetWorkflowExecutionHistory' and isinstance(err, EntityNotExistsError): |
Member
There was a problem hiding this comment.
tbh not sure I understand what would trigger this besides replication lag races?
Member
Author
There was a problem hiding this comment.
Yeah that's exactly what it's for. If you Start a Workflow and want to poll until it completes you have to do this.
shijiesheng
reviewed
Sep 8, 2025
cadence/_internal/rpc/retry.py
Outdated
| except CadenceError as e: | ||
| err = e | ||
| attempts = 1 | ||
| while is_retryable(err, client_call_details): |
Member
There was a problem hiding this comment.
nit: simplify using do while loop by using break/continue
cadence/_internal/rpc/retry.py
Outdated
|
|
||
| def is_retryable(err: CadenceError, call_details: ClientCallDetails) -> bool: | ||
| # Handle requests to the passive side, matching the Go and Java Clients | ||
| if call_details.method == b'/uber.cadence.api.v1.WorkflowAPI/GetWorkflowExecutionHistory' and isinstance(err, EntityNotExistsError): |
Member
There was a problem hiding this comment.
I tried to use DESCRIPTOR but that's too complicated. Maybe at least add a variable to replace the string here?
shijiesheng
approved these changes
Sep 9, 2025
Add RPC retries via an interceptor, with exponential backoff matching both the Go and Java clients. The approach here differs from both however in two main ways: 1. Adding these via interceptor makes it implicit, while both the Go and Java client require it to be explicit. 2. The specific requests to retry are based on GRPC error codes, rather than explicitly listing non-retryable errors and retrying everything by default. This seems like a more sustainable approach, since nearly every error type is non-retryable. A newly introduced error type would require a client update to mark it non-retryable before it could safely be used. Any time the python client doesn't recognize an error it gets mapped to just CadenceError, so new errors can safely be added.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add RPC retries via an interceptor, with exponential backoff matching both the Go and Java clients. The approach here differs from both however in two main ways:
What changed?
Why?
How did you test it?
Potential risks
Release notes
Documentation Changes