fix(python): Retry connections on HTTP/2 API transports - #1425
Conversation
🦋 Changeset detectedLatest commit: 8ffd738 The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
PR SummaryLow Risk Overview A new Reviewed by Cursor Bugbot for commit 8ffd738. Bugbot is set up for automated code reviews on this repo. Configure here. |
Package ArtifactsBuilt from 7301a24. Download artifacts from this workflow run. JS SDK ( npm install ./e2b-2.29.2-wj-fix-retries.0.tgzCLI ( npm install ./e2b-cli-2.11.2-wj-fix-retries.0.tgzPython SDK ( pip install ./e2b-2.28.2+wj.fix.retries-py3-none-any.whl |
Pass retries to the httpx HTTP/2 transports used for the E2B API and envd clients (sync and async) so transient connect failures are retried. Defaults to 3 and is configurable via the E2B_CONNECTION_RETRIES environment variable, following the same pattern as the other connection pool settings. httpx only retries connection establishment errors, so this is safe for non-idempotent requests.
…ents Main moved the sync SDK to per-thread RPC/HTTP clients (#1422, #1425). Rebuild the sandbox health probe on top of that model: sync Commands and Pty now lazily create a thread-local envd HTTP client (mirroring Filesystem) and expose _check_health as a method resolving the calling thread's client, instead of capturing a single shared httpx.Client at construction time. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
3 participants
No description provided.