transport: log network reconnects with same peer process #128415

schase-es · 2025-05-24T01:19:53Z

ClusterConnectionManager now caches the previous ephemeralId (created on process-start) of peer nodes on disconnect in a connection history table. On reconnect, when a peer has the same ephemeralId as it did previously, this is logged to indicate a network failure. The connectionHistory is trimmed to the current set of peers by NodeConnectionsService.

elasticsearchmachine · 2025-05-24T01:20:17Z

Pinging @elastic/es-distributed-coordination (Team:Distributed Coordination)

schase-es · 2025-05-24T01:20:33Z

I wasn't able to find a way to test the ClusterConnectionManager's connectionHistory table when integrated through the NodeConnectionsService.

nicktindall

Looking good, just a few questions and minor comments.

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

nicktindall · 2025-05-26T01:25:18Z

server/src/main/java/org/elasticsearch/transport/ConnectionManager.java

+    /**
+     * Keep the connection history for the nodes listed
+     */
+    void retainConnectionHistory(List<DiscoveryNode> nodes);


In the javadoc I think we should mention that we discard history for nodes not in the list? If you know the Set API then it's suggested by the name retain, but if you don't it might not be obvious.

nicktindall · 2025-05-26T01:33:49Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

@@ -120,6 +122,7 @@ public void connectToNodes(DiscoveryNodes discoveryNodes, Runnable onCompletion)
                        runnables.add(connectionTarget.connect(null));
                    }
                }
+                transportService.retainConnectionHistory(nodes);


We might be able to use DiscoveryNodes#getAllNodes() rather than building up an auxiliary collection, that might be marginally more efficient? Set#retainAll seems to take a Collection, but we'd need to change the ConnectionManager#retainConnectionHistory interface to accommodate.

Do we need a separate collection here at all? We could just pass discoveryNodes around I think.

But also, really this is cleaning out the nodes about which we no longer care, so I think we should be doing this in disconnectFromNodesExcept instead.

Nick raised an important point about the race between the connection history table and the close callback.

A connection's close callback will always put an entry in the history table. If this close is a consequence of a cluster state change and disconnect in NodeConnectionsService, then it will add a node history right after it's supposed to be cleaned out.

Cleaning out the node history table whenever we disconnect from some nodes or connect to some new nodes works fine, but it means the history table will always lag a version behind, in what it's holding onto.

I came up with a concurrency scheme that works for keeping the node history current in NodeConnectionsService, but it's more complicated.

nicktindall · 2025-05-26T02:07:31Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

+                                    public void onFailure(Exception e) {
+                                        final NodeConnectionHistory hist = new NodeConnectionHistory(node.getEphemeralId(), e);
+                                        nodeHistory.put(conn.getNode().getId(), hist);
+                                    }


Do we want to store the connection history even when conn.hasReferences() == false ? I'm not 100% familiar with this code, but I wonder if we might get the occasional ungraceful disconnect after we've released all our references?

I guess in that case we would eventually discard the entry via retainConnectionHistory anyway.

Do we need to be careful with the timing of calls to retainConnectionHistory versus the these close handlers firing? I guess any entries that are added after a purge would not survive subsequent purges.

nicktindall · 2025-05-26T02:14:33Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

+                                                node.descriptionWithoutAttributes(),
+                                                e,
+                                                ReferenceDocs.NETWORK_DISCONNECT_TROUBLESHOOTING
+                                            );


It looks like previously we would only have logged at debug level in this scenario? unless I'm reading it wrong. I'm not sure how interesting this case is (as we were disconnecting from the node anyway)?

nicktindall · 2025-05-26T02:23:10Z

server/src/test/java/org/elasticsearch/transport/ClusterConnectionManagerTests.java

+            assertTrue("recent disconnects should be listed", connectionManager.connectionHistorySize() == 2);
+
+            connectionManager.retainConnectionHistory(Collections.emptyList());
+            assertTrue("connection history should be emptied", connectionManager.connectionHistorySize() == 0);


I wonder if it would be better to expose a read-only copy of the map for testing this, that would allow us to assert that the correct IDs were present?

DaveCTurner

I think ClusterConnectionManager isn't quite the right place to do this - the job of this connection manager is to look after all node-to-node connections including ones used for discovery and remote cluster connections too. There are situations where we might close and re-establish these kinds of connection without either end restarting without that being a problem worthy of logging.

NodeConnectionsService is the class that knows about connections to nodes in the cluster. I'd rather we implemented the logging about unexpected reconnects there. That does raise some difficulties about how to expose the exception that closed the connection, if such an exception exists. I did say that this bit would be tricky 😁 Nonetheless I'd rather we got the logging to happen in the right place first and then we can think about the plumbing needed to achieve this extra detail.

DaveCTurner · 2025-05-27T07:09:15Z

...tty4/src/internalClusterTest/java/org/elasticsearch/transport/netty4/ESLoggingHandlerIT.java

+        value = "org.elasticsearch.transport.ClusterConnectionManager:WARN",
+        reason = "to ensure we log cluster manager disconnect events on WARN level"
+    )
+    public void testExceptionalDisconnectLoggingInClusterConnectionManager() throws Exception {


Could we put this into its own test suite? This suite is supposed to be about ESLoggingHandler which is unrelated to the logging in ClusterConnectionManager. I think this test should work fine in the :server test suite, no need to hide it in the transport-netty4 module.

Also could you open a separate PR to move testConnectionLogging and testExceptionalDisconnectLogging out of this test suite - they're testing the logging in TcpTransport which is similarly unrelated to ESLoggingHandler. IIRC they were added here for historical reasons, but these days we use the Netty transport everywhere so these should work in :server too.

DaveCTurner · 2025-05-27T07:12:23Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

@@ -120,6 +122,7 @@ public void connectToNodes(DiscoveryNodes discoveryNodes, Runnable onCompletion)
                        runnables.add(connectionTarget.connect(null));
                    }
                }
+                transportService.retainConnectionHistory(nodes);


Do we need a separate collection here at all? We could just pass discoveryNodes around I think.

But also, really this is cleaning out the nodes about which we no longer care, so I think we should be doing this in disconnectFromNodesExcept instead.

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

DaveCTurner · 2025-05-27T07:25:44Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

+                            NodeConnectionHistory hist = nodeHistory.remove(connNode.getId());
+                            if (hist != null && hist.ephemeralId.equals(connNode.getEphemeralId())) {


Could we extract this to a separate method rather than adding to this already over-long and over-nested code directly?

Also I'd rather use nodeConnectionHistory instead of hist. Abbreviated variable names are a hindrance to readers, particularly if they don't have English as a first language, and there's no disadvantage to using the full type name here.

(nit: also it can be final)

DaveCTurner · 2025-05-27T07:31:11Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

+                                if (hist.disconnectCause != null) {
+                                    logger.warn(
+                                        () -> format(
+                                            "transport connection reopened to node with same ephemeralId [%s], close exception:",


Users don't really know what ephemeralId is so I think will find this message confusing. Could we say something like reopened transport connection to node [%s] which disconnected exceptionally [%s/%dms] ago but did not restart, so the disconnection is unexpected? NB also tracking the disconnection duration here.

Similarly disconnected gracefully in the other branch.

Also can we link ReferenceDocs.NETWORK_DISCONNECT_TROUBLESHOOTING?

DaveCTurner · 2025-05-27T07:36:49Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

-                                        // that's a bug.
-                                    } else {
-                                        logger.debug("closing unused transport connection to [{}]", node);
+                                conn.addCloseListener(new ActionListener<Void>() {


nit: reduce duplication a bit here:

conn.addCloseListener(new ActionListener<>() { @Override public void onResponse(Void ignored) { addNewNodeConnectionHistory(null); } @Override public void onFailure(Exception e) { addNewNodeConnectionHistory(e); } private void addNewNodeConnectionHistory(@Nullable Exception e) { nodeHistory.put(node.getId(), new NodeConnectionHistory(node.getEphemeralId(), e)); } });

Also consider extracting this out to the top level to try and keep this method length/nesting depth from getting too much further out of hand.

schase-es · 2025-05-27T21:55:35Z

Thanks for the feedback everyone. It looks like I can repurpose the TransportConnectionListener interface to get the connect/disconnect events in node connections service, and use the connection close listener to get any exception out.

- moved test out of ESLoggingHandlerIt into a separate ClusterConnectionManagerIntegTests file - moved connection history into NodeConnectionsService, and adopted a consistency scheme - rewrote re-connection log message to include duration - changed log level of local disconnect with exception to debug

DaveCTurner · 2025-05-29T07:40:13Z

server/src/main/java/org/elasticsearch/transport/ClusterConnectionManager.java

+                                            logger.warn(
+                                                """
+                                                    transport connection to [{}] closed by remote with exception [{}]; \
+                                                    if unexpected, see [{}] for troubleshooting guidance""",


I think this isn't guaranteed to be a WARN worthy event - if the node shut down then we might get a Connection reset or similar but that's not something that needs action, and we do log those exceptions elsewhere. On reflection I'd rather leave the logging in ClusterConnectionManager alone in this PR and just look at the new logs from the NodeConnectionsService.

DaveCTurner · 2025-05-29T07:41:40Z

server/src/test/java/org/elasticsearch/transport/ClusterConnectionManagerIntegTests.java

+import org.elasticsearch.test.junit.annotations.TestLogging;
+
+@ESIntegTestCase.ClusterScope(numDataNodes = 2, scope = ESIntegTestCase.Scope.TEST)
+public class ClusterConnectionManagerIntegTests extends ESIntegTestCase {


nit: ESIntegTestCase tests should have names ending in IT and be in the internalClusterTest source set. But as mentioned in my previous comment we probably don't want to change this here.

DaveCTurner · 2025-05-29T07:42:41Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

@@ -347,4 +357,113 @@ public String toString() {
            }
        }
    }
+
+    private class ConnectionHistory {


Yeah I like the look of this. Maybe ConnectionHistory implements TransportConnectionListener rather than having another layer of indirection?

Also this needs to be covered in NodeConnectionsServiceTests.

DaveCTurner · 2025-05-29T07:45:27Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+         * Each node in the cluster always has a nodeHistory entry that is either the dummy value or a connection history record. This
+         * allows node disconnect callbacks to discard their entry if the disconnect occurred because of a change in cluster state.
+         */
+        private final NodeConnectionHistory dummy = new NodeConnectionHistory("", 0, null);


Can be static I think, it's a global constant. We tend to name global constants in SHOUTY_SNAKE_CASE reflecting their meaning, so here I'd suggest CONNECTED or CONNECTED_MARKER or something like that. This way you get to say nodeConnectionHistory != CONNECTED_MARKER below which makes it clearer to the reader what this predicate means.

nit: also looks like the javadoc is for the nodeHistory field

DaveCTurner · 2025-05-29T07:46:19Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+                                    "reopened transport connection to node [%s] "
+                                        + "which disconnected exceptionally [%dms] ago but did not "
+                                        + "restart, so the disconnection is unexpected; "
+                                        + "if unexpected, see [{}] for troubleshooting guidance",


No need for if unexpected here, I think the point is that this situation is always unexpected.

DaveCTurner · 2025-05-29T07:46:57Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+                                        + "restart, so the disconnection is unexpected; "
+                                        + "if unexpected, see [{}] for troubleshooting guidance",
+                                    node.descriptionWithoutAttributes(),
+                                    nodeConnectionHistory.disconnectTime,


This'll show the absolute disconnect time in milliseconds (i.e. since 1970) whereas I think we want to see the duration between the disconnect and the current time.

schase-es · 2025-05-29T21:50:02Z

Thanks for the feedback David -- this was definitely a light pass on everything other than the concurrency scheme, and I wanted to get notes on it before adding complete testing and getting everything else just right. In hindsight, I was probably better off not trying to address everything else at the same time instead of committing first-draft versions.

nicktindall · 2025-05-30T06:20:02Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+
+        void reserveConnectionHistoryForNodes(DiscoveryNodes nodes) {
+            for (DiscoveryNode node : nodes) {
+                nodeHistory.put(node.getId(), dummy);


This might need to be putIfAbsent so we don't over-write any actual current NodeConnectionHistory entries right?

I'm not sure. My read was these two calls would come from cluster state changing to add or remove nodes from this table. Inclusion is controlled by these calls, which unconditionally add or remove entries. The close callback has to be careful to check if it has an entry that's valid: this protects against long-running callbacks inserting garbage into the table.

The DiscoveryNodes passed to connectToNodes contains all the nodes in the cluster, including any existing ones, so if there's a node which already exists in the cluster, and is currently disconnected, then it will have an entry in nodeHistory which isn't dummy that this line will overwrite on any cluster state update. So yeah I think putIfAbsent is what we want here.

I get it now -- for whatever reason, I thought it was passing in the deltas, but it's obvious from connectToNodes that the node connections service is doing that calculation

nicktindall · 2025-05-30T06:21:31Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+            });
+        }
+
+        void reserveConnectionHistoryForNodes(DiscoveryNodes nodes) {


nit: I wonder if this should be called something like startTrackingConnectionHistory (and the other method stop...), the "reserving" language seems like an implementation detail leaking?

I do like the implementation though, nice approach to fixing the race.

DaveCTurner · 2025-05-30T07:37:13Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+
+        void reserveConnectionHistoryForNodes(DiscoveryNodes nodes) {
+            for (DiscoveryNode node : nodes) {
+                nodeHistory.put(node.getId(), dummy);


The DiscoveryNodes passed to connectToNodes contains all the nodes in the cluster, including any existing ones, so if there's a node which already exists in the cluster, and is currently disconnected, then it will have an entry in nodeHistory which isn't dummy that this line will overwrite on any cluster state update. So yeah I think putIfAbsent is what we want here.

DaveCTurner · 2025-05-30T07:46:50Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+                    NodeConnectionHistory nodeConnectionHistory = nodeHistory.get(node.getId());
+                    if (nodeConnectionHistory != null) {
+                        nodeHistory.replace(node.getId(), nodeConnectionHistory, dummy);
+                    }


This looks a little racy, although in practice I think it's fine because ClusterConnectionManager protects against opening multiple connections to the same node concurrently. Still, if we did all this (including the logging) within a nodeHistory.compute(node.getId, ...) then there'd obviously be no races.

DaveCTurner · 2025-05-30T07:53:17Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+        void removeConnectionHistoryForNodes(Set<DiscoveryNode> nodes) {
+            final int startSize = nodeHistory.size();
+            for (DiscoveryNode node : nodes) {
+                nodeHistory.remove(node.getId());


There's kind of an implicit invariant here that org.elasticsearch.cluster.NodeConnectionsService.ConnectionHistory#nodeHistory and org.elasticsearch.cluster.NodeConnectionsService#targetsByNode have the same keys. At the very least we should be able to assert this. I also wonder if we should be calling nodeHistory.retainAll() to make it super-clear that we are keeping these keysets aligned.

But then that got me thinking, maybe we should be tracking the connection history of each target node in ConnectionTarget rather than trying to maintain two parallel maps. Could that work?

This is a great idea... ConnectionTarget has exactly the lifecycle needed. I think because I moved it from elsewhere and am having a rough week over here, this didn't occur to me.

DaveCTurner · 2025-05-30T09:00:07Z

server/src/main/java/org/elasticsearch/cluster/NodeConnectionsService.java

+                }
+
+                @Override
+                public void onNodeDisconnected(DiscoveryNode node, Transport.Connection connection) {


I just spotted we're already executing this in a close-listener, but one that runs under ActionListener.running(...) so it drops the exception. I think it'd be nicer to adjust this callback to take a @Nullable Exception e parameter rather than having to add a second close listener just to pick up the exception as done here.

- consolidated ConnectionHistory into ConnectionTarget, protected with the service mutex - added logging test for reconnection with and without exception - grew TransportConnectionListener onNodeDisconnected to include a nullable exception - reverted ClusterConnectionManager tests and logging

schase-es · 2025-05-31T01:05:21Z

I think this addresses everything so far. The test doesn't check on the internals of ConnectionTarget -- but no other tests do. I do have a more complete tests I can adapt.

One question/issue that came up earlier in the discussion of this PR was around the lifecycle of DiscoveryNode. Because:

the ephemeralId field is immutable and used for hashing and equality
and the ephemeralId is based on information received after a connection is established
it's not possible for the discovery node view that the NodeConnectionsService has when it establishes the connection to have the correct ephemeralId as the node view after the connection has been established and handshaking has occurred. (At least, this was the hypothesis -- I haven't been able to track down where the ephemeralId gets set).

So far, I've been careful to use the service's view for storage and retrieval, and the connection's view for comparison. It's particularly awkward in this implementation. I am wondering if this makes sense to David: whether this idea/concern is valid, and whether it works out if this is true.

I am also hoping to do a real-world test, or have something with more complete integration.

schase-es requested review from nicktindall and DaveCTurner May 24, 2025 01:19

schase-es added >non-issue :Distributed Coordination/Network Http and internode communication implementations labels May 24, 2025

elasticsearchmachine added the Team:Distributed Coordination Meta label for Distributed Coordination team label May 24, 2025

elasticsearchmachine added the v9.1.0 label May 24, 2025

[CI] Auto commit changes from spotless

bd688bb

nicktindall reviewed May 26, 2025

View reviewed changes

DaveCTurner reviewed May 27, 2025

View reviewed changes

DaveCTurner reviewed May 29, 2025

View reviewed changes

nicktindall reviewed May 30, 2025

View reviewed changes

DaveCTurner reviewed May 30, 2025

View reviewed changes

		NodeConnectionHistory hist = nodeHistory.remove(connNode.getId());
		if (hist != null && hist.ephemeralId.equals(connNode.getEphemeralId())) {

transport: log network reconnects with same peer process #128415

Are you sure you want to change the base?

transport: log network reconnects with same peer process #128415

Conversation

schase-es commented May 24, 2025

Uh oh!

elasticsearchmachine commented May 24, 2025

Uh oh!

schase-es commented May 24, 2025

Uh oh!

nicktindall left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

schase-es May 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DaveCTurner left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

schase-es commented May 27, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

schase-es commented May 29, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

nicktindall May 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

schase-es May 29, 2025 •

edited

Loading

nicktindall May 30, 2025 •

edited

Loading