rpuch commented on code in PR #7964:
URL: https://github.com/apache/ignite-3/pull/7964#discussion_r3073757418
##########
modules/network/src/main/java/org/apache/ignite/internal/network/DefaultMessagingService.java:
##########
@@ -465,12 +504,139 @@ private CompletableFuture<Void> sendViaNetwork(
return sender.send(
new OutNetworkObject(message, descriptors),
- () -> triggerChannelCreation(nodeId, type, addr)
+ () -> triggerChannelCreation(nodeId, type, addr,
strictIdCheck, retryStrategy, channelTimeoutMillis)
);
});
}
- private void maybeLogHandshakeError(Throwable ex, UUID nodeId, ChannelType
type, InetSocketAddress addr) {
+ private OrderingFuture<NettySender> openChannelWithRetries(
+ @Nullable UUID nodeId,
+ ChannelType type,
+ InetSocketAddress addr,
+ boolean strictIdCheck,
+ TerminalIoErrorRetryStrategy retryStrategy,
+ long timeoutMillis
+ ) {
+ Long deadlineMillis = timeoutMillis == NO_TIMEOUT ||
!timeoutDefined(timeoutMillis) ? null
+ : coarseCurrentTimeMillis() + timeoutMillis;
+ if (deadlineMillis != null && deadlineMillis < 0) {
Review Comment:
It is about the case when a caller just does `invoke(..., Long.MAX_VALUE)`
to get a 'no timeout' behavior. I think we already had such code somewhere
about Raft snapshots installation.
This code is only executed when creating a new TCP connection, which is not
a frequent operation. Why do you believe boxing is a problem here?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
