wu-sheng commented on code in PR #777:
URL: https://github.com/apache/skywalking-java/pull/777#discussion_r2516428523
##########
apm-sniffer/apm-agent-core/src/main/java/org/apache/skywalking/apm/agent/core/remote/GRPCChannelManager.java:
##########
@@ -184,21 +184,65 @@ public Channel getChannel() {
*/
public void reportError(Throwable throwable) {
if (isNetworkError(throwable)) {
- reconnect = true;
- notify(GRPCChannelStatus.DISCONNECT);
+ triggerReconnect();
}
}
private void notify(GRPCChannelStatus status) {
- for (GRPCChannelListener listener : listeners) {
- try {
- listener.statusChanged(status);
- } catch (Throwable t) {
- LOGGER.error(t, "Fail to notify {} about channel connected.",
listener.getClass().getName());
+ synchronized (listeners) {
+ for (GRPCChannelListener listener : listeners) {
+ try {
+ listener.statusChanged(status);
+ } catch (Throwable t) {
+ LOGGER.error(t, "Fail to notify {} about channel
connected.", listener.getClass().getName());
+ }
}
}
}
+ /**
+ * Create a new gRPC channel to the specified server and reset connection
state.
+ */
+ private void createNewChannel(String host, int port) throws Exception {
+ if (managedChannel != null) {
+ managedChannel.shutdownNow();
+ }
+
+ managedChannel = GRPCChannel.newBuilder(host, port)
+ .addManagedChannelBuilder(new
StandardChannelBuilder())
+ .addManagedChannelBuilder(new
TLSChannelBuilder())
+ .addChannelDecorator(new
AgentIDDecorator())
+ .addChannelDecorator(new
AuthenticationDecorator())
+ .build();
+
+ // Reset reconnectCount after actually rebuilding the channel
+ reconnectCount = 0;
+ notifyConnected();
+ }
+
+ /**
+ * Trigger reconnection by setting reconnect flag and notifying listeners.
+ */
+ private void triggerReconnect() {
+ synchronized (statusLock) {
+ reconnect = true;
+ notify(GRPCChannelStatus.DISCONNECT);
+ }
+ }
+
+ /**
+ * Notify listeners that connection is established without resetting
reconnectCount.
+ * This is used when the channel appears connected but we want to keep
monitoring
+ * reconnect attempts in case it's a false positive (half-open connection).
+ */
+ private void notifyConnected() {
+ synchronized (statusLock) {
+ // Don't reset reconnectCount - connection might still be half-open
Review Comment:
Isn't a way you can determine the server is reachable? I am a little
confused. Still no `TRANSIENT_FAILURE` status check in your codes.
--
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]
