Github user gemmellr commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/390#discussion_r52922134
--- Diff:
artemis-protocols/artemis-proton-plug/src/main/java/org/proton/plug/context/AbstractConnectionContext.java
---
@@ -40,28 +44,32 @@
public abstract class AbstractConnectionContext extends
ProtonInitializable implements AMQPConnectionContext {
+ public static final Symbol CONNECTION_OPEN_FAILED =
Symbol.valueOf("amqp:connection-establishment-failed");
+
protected ProtonHandler handler = ProtonHandler.Factory.create();
protected AMQPConnectionCallback connectionCallback;
+ private final ScheduledExecutorService scheduledPool;
private final Map<Session, AbstractProtonSessionContext> sessions = new
ConcurrentHashMap<>();
protected LocalListener listener = new LocalListener();
- public AbstractConnectionContext(AMQPConnectionCallback
connectionCallback) {
- this(connectionCallback, DEFAULT_IDLE_TIMEOUT,
DEFAULT_MAX_FRAME_SIZE, DEFAULT_CHANNEL_MAX);
+ public AbstractConnectionContext(AMQPConnectionCallback
connectionCallback, ScheduledExecutorService scheduledPool) {
+ this(connectionCallback, DEFAULT_IDLE_TIMEOUT,
DEFAULT_MAX_FRAME_SIZE, DEFAULT_CHANNEL_MAX, scheduledPool);
}
public AbstractConnectionContext(AMQPConnectionCallback
connectionCallback,
int idleTimeout,
int maxFrameSize,
- int channelMax) {
+ int channelMax,
+ ScheduledExecutorService
scheduledPool) {
this.connectionCallback = connectionCallback;
+ this.scheduledPool = scheduledPool;
connectionCallback.setConnection(this);
Transport transport = handler.getTransport();
if (idleTimeout > 0) {
- transport.setIdleTimeout(idleTimeout);
- transport.tick(idleTimeout / 2);
+ transport.setIdleTimeout(idleTimeout / 2);
--- End diff --
Meant to comment on this before...did you mean to half the value passed
here, it wasn't halved before?
Proton will half the given value before advertising it to the peer since
the AMQP spec suggests you should advertise half your actual timeout (proton
also pessimistically assumes on the receiving side that the peer did not half
it, and so halves it again to ensure it sends more often than the actual
timeout).
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
