fix: retry + 0.5s cooldown on MWP POST after R2 phase to fix SSLV3_ALERT_BAD_RECORD_MAC
After 200+ rapid R2 TLS connections, the TCP/TLS stack enters a transiently unstable state that triggers BAD_RECORD_MAC on the next CF connection. The 0.5s sleep lets the kernel drain sockets and CF edge state stabilize. The retry loop (up to 3 attempts, 1.5^n backoff) handles any residual transient failure, mirroring the identical fix already in _presigned_put.
sha256:446fd9127a1ec36bc7c71cb005d734046dd4cc8bcb3cd2c54c14b99eb66aab78
sha
sha256:9cfda7aaa1a2495c79ac4dc167c6a74d45677aaf375ac3deb4f118af6cbfb4ae
snapshot
0 comments
To add a comment, use the Muse CLI:
muse hub commit comment sha256:446fd9127a1ec36bc7c71cb005d734046dd4cc8bcb3cd2c54c14b99eb66aab78 --body "your comment"
No comments yet. Be the first to start the discussion.