CLDSRV-968: Expire idle keep-alive sockets in the functional test client - #6247
Conversation
Hello tcarmet,My role is to assist you with the merge of this Available options
Available commands
Status report is not available. |
Incorrect fix versionThe
Considering where you are trying to merge, I ignored possible hotfix versions and I expected to find:
Please check the |
Request integration branchesWaiting for integration branch creation to be requested by the user. To request integration branches, please comment on this pull request with the following command: Alternatively, the |
|
/create_integration_branches |
Integration data createdI have created the integration data for the additional destination branches.
The following branches will NOT be impacted:
You can set option The following options are set: create_integration_branches |
Waiting for approvalThe following approvals are needed before I can proceed with the merge:
The following options are set: create_integration_branches |
Incorrect fix versionThe
Considering where you are trying to merge, I ignored possible hotfix versions and I expected to find:
Please check the The following options are set: create_integration_branches |
The SDK agents in the aws-node-sdk test client enable keepAlive without an idle
timeout, so pooled sockets are never expired client-side. Node's HTTP server
closes idle keep-alive connections after 5s and advertises `Keep-Alive:
timeout=5`, but the agent only honours that hint when it already has a non-zero
timeout of its own. Without one, a burst of requests issued after an idle period
can be written to connections the server has already closed, failing with
ECONNRESET ("socket hang up").
Set a 4s agent idle timeout so the client retires sockets before the server
does. This only destroys sockets sitting in the free pool; in-flight requests
stay governed by requestTimeout.
0065a20 to
7346130
Compare
|
/reset |
Reset completeI have successfully deleted this pull request's integration branches. The following options are set: create_integration_branches |
Waiting for approvalThe following approvals are needed before I can proceed with the merge:
The following options are set: create_integration_branches |
|
/approve |
Build failedThe build for commit did not succeed in branch w/9.4/bugfix/CLDSRV-968-test-client-idle-socket-timeout The following options are set: approve, create_integration_branches |
Build failedThe build for commit did not succeed in branch bugfix/CLDSRV-968-test-client-idle-socket-timeout The following options are set: approve, create_integration_branches |
❌ 1 Tests Failed:
View the top 1 failed test(s) by shortest run time
View the full list of 1 ❄️ flaky test(s)
To view more test analytics, go to the Test Analytics Dashboard |
Build failedThe build for commit did not succeed in branch w/9.4/bugfix/CLDSRV-968-test-client-idle-socket-timeout The following options are set: approve, create_integration_branches |
Build failedThe build for commit did not succeed in branch bugfix/CLDSRV-968-test-client-idle-socket-timeout The following options are set: approve, create_integration_branches |
In the queueThe changeset has received all authorizations and has been added to the The changeset will be merged in:
The following branches will NOT be impacted:
This pull request does not target the following hotfix branch(es) so they
There is no action required on your side. You will be notified here once IMPORTANT Please do not attempt to modify this pull request.
If you need this pull request to be removed from the queue, please contact a The following options are set: approve, create_integration_branches |
Queue build failedThe corresponding build for the queue failed:
Remove the pull request from the queue
|
|
I have successfully merged the changeset of this pull request
The following branches have NOT changed:
Please check the status of the associated issue CLDSRV-968. Goodbye tcarmet. |
The functional test client keeps pooled keep-alive connections indefinitely, so after an idle gap it can send a request over a connection the server has already closed and fail with a spurious connection reset. Because the client also runs with retries disabled, one reset is enough to fail a hook and take a whole group of tests down with it, which surfaces as intermittent CI red on slower full-deployment environments where such idle gaps actually occur.
This makes the client expire its own idle connections before the server does, so the reuse race cannot arise.