-
Notifications
You must be signed in to change notification settings - Fork 4.8k
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
Windows: Fix ssl_socket_test #13264
Windows: Fix ssl_socket_test #13264
Conversation
- SmallReadsIntoSameSlice test: use write method on connection instead of transport socket so SSL_write failures will be properly retried, also otherwise empty writes from connection will trigger failed assertions in buffer::linearize - When SSL shutdown is not appropriate, perform basic socket shutdown; relevant when client certificate is invalid and connection should be dropped by server, however we still need to ensure TLS alerts are able to be read by client, shutdown on socket ensures this can happen - Open files for test in text mode to take care of line ending issues Signed-off-by: Sunjay Bhatia <[email protected]>
cc @envoyproxy/windows-dev |
/assign @PiotrSikora |
Need to make binary mode open optional/refactor some tests Also some of the ssl_socket_tests seem to fail in RBE and not locally, another thing to dig into /wait |
Signed-off-by: Sunjay Bhatia <[email protected]>
Looks good besides a small fix. Thanks a lot for getting this one to work! |
Signed-off-by: Sunjay Bhatia <[email protected]>
Signed-off-by: William A Rowe Jr <[email protected]>
I believe the openssl fork errors are either caused by the Windows update service, or MSYS2 issues, hopefully addressed by one of: envoyproxy/envoy-build-tools#101 or envoyproxy/envoy-build-tools#100 |
See #13276 Looks like envoyproxy/envoy-build-tools#100 is required to solve this issue in the build container/RBE |
/wait |
/wait waiting on #13367 |
Failing due to same reason this test is being marked as flaky for the time being: #13407 |
/wait still waiting on resolving the openssl msys2 fork issue, may continue with just checking in certs etc. |
Blocked on whichever #13702 or envoyproxy/envoy-build-tools#107 is chosen as the solution |
…ket-test Signed-off-by: Sunjay Bhatia <[email protected]>
unblocked by #13702 |
/retest |
Retrying Azure Pipelines. |
@PiotrSikora this should be ready to go again, macos CI failure looks like one of the systematic ones we've been having generally |
…ket-test Signed-off-by: William A Rowe Jr <[email protected]>
merged master again to hopefully get the non-windows flakes out, @PiotrSikora have you had a chance to take a look at this yet? |
* master: (83 commits) tls: Typesafe tls slots (envoyproxy#13789) docs(example): Correct URL for caching example page (envoyproxy#13810) [fuzz] Made health check fuzz more efficient (envoyproxy#13747) rtds: properly scope rtds stats (envoyproxy#13764) http: fixing a bug with IPv6 hosts (envoyproxy#13798) connection: Remember transport socket read resumption requests and replay them when re-enabling read. (envoyproxy#13772) network: adding some accessors for ALPN work. (envoyproxy#13785) docs: added a step about how to handle platform specific extensions (envoyproxy#13759) Fix identation in ip transparency code snippet (envoyproxy#13743) wasm: enable WAVM's stack unwinding feature (envoyproxy#13792) log: set route name for direct response (envoyproxy#13683) Use nghttp2 as external dependsncy in protocol_constraints_lib (envoyproxy#13763) [Windows] Update windows dev docs (envoyproxy#13741) cel: patch thread safety issue (envoyproxy#13739) Windows: Fix ssl_socket_test (envoyproxy#13264) apple dns: add fake api test suite (envoyproxy#13780) overload: scale selected timers in response to load (envoyproxy#13475) examples: Add dynamic configuration (control plane) sandbox (envoyproxy#13746) Removed exception in getResponseStatus() (envoyproxy#13314) network: add timeout for transport connect (envoyproxy#13610) ... Signed-off-by: Michael Puncel <[email protected]>
Commit Message:
Windows: Fix ssl_socket_test
of transport socket so SSL_write failures will be properly retried,
also otherwise empty writes from connection will trigger failed
assertions in buffer::linearize
relevant when client certificate is invalid and connection should be
dropped by server, however we still need to ensure TLS alerts are
able to be read by client, shutdown on socket ensures this can happen
\r
with""
when comparing test fixtures to results to ensure lineendings are ignored
Risk Level: Low
Testing: Locally on Windows
Docs Changes: N/A
Release Notes: N/A
Fixes #13191