Skip to content

Fix invalid parameter issue when force to close endpoint #25

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.

Sign up for GitHub

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

Open
brianplus wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix invalid parameter issue when force to close endpoint #25

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 8 additions & 1 deletion src/main/scala/org/apache/spark/shuffle/ucx/UcxWorkerWrapper.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -188,7 +188,14 @@ case class UcxWorkerWrapper(worker: UcpWorker, transport: UcxShuffleTransport, i
def connectByWorkerAddress(executorId: transport.ExecutorId, workerAddress: ByteBuffer): Unit = {
logDebug(s"Worker $this connecting back to $executorId by worker address")
val ep = worker.newEndpoint(new UcpEndpointParams().setName(s"Server connection to $executorId")
.setUcpAddress(workerAddress))
.setUcpAddress(workerAddress)
.setPeerErrorHandlingMode()
Copy link
Copy Markdown
Member

@petro-rudenko petro-rudenko Mar 10, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@yosefe does peerErrorHandling mode works with connect by worker address?

Copy link
Copy Markdown

@leibin2014 leibin2014 Mar 23, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@yosefe does peerErrorHandling mode works with connect by worker address?

@yosefe Can you please review? Thanks!

Copy link
Copy Markdown
Author

@brianplus brianplus May 12, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In UCX,

ucs_status_ptr_t ucp_ep_close_nbx(ucp_ep_h ep, const ucp_request_param_t *param)
{
    ucp_worker_h  worker = ep->worker;
    void          *request = NULL;
    ucp_request_t *close_req;

    if ((ucp_request_param_flags(param) & UCP_EP_CLOSE_FLAG_FORCE) &&
        (ucp_ep_config(ep)->key.err_mode != UCP_ERR_HANDLING_MODE_PEER)) {
        return UCS_STATUS_PTR(UCS_ERR_INVALID_PARAM);
    }

when force to close the ep, it will also check the flag UCP_ERR_HANDLING_MODE_PEER setting.

Copy link
Copy Markdown

@leibin2014 leibin2014 Oct 24, 2023

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@gleon99 @yosefe Can you pls help to review and merge this PR? Thanks!

.setErrorHandler(new UcpEndpointErrorHandler() {
override def onError(ep: UcpEndpoint, status: Int, errorMsg: String): Unit = {
logError(s"Endpoint to $executorId got an error: $errorMsg")
connections.remove(executorId)
}
}))
connections.put(executorId, ep)
}

Expand Down