Open sasbury opened 5 years ago
could be that we are too agressive closing the stan connection or at least "forgetting" about it somehow so we try to recreate when we shouldn't
set the connectionlost handler and only try to reconnect when you get notification from that cb, not the low level NATS disconnect handler
Will this get fixed?
ug 21 00:26:27 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:27.634026 [INF] error restarting streaming connection, will retry in 5000 milliseconds%!(EXTRA string=stan: clientID already registered) Aug 21 00:26:32 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:32.634170 [INF] trying to reconnect to nats streaming Aug 21 00:26:32 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:32.634198 [INF] connecting to NATS streaming with configuration stan-eu-p1, cluster id is rg-cluster-eu-central-1 Aug 21 00:26:32 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:32.635578 [INF] error restarting streaming connection, will retry in 5000 milliseconds%!(EXTRA string=stan: clientID already registered) Aug 21 00:26:37 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:37.635728 [INF] trying to reconnect to nats streaming Aug 21 00:26:37 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:37.635755 [INF] connecting to NATS streaming with configuration stan-eu-p1, cluster id is rg-cluster-eu-central-1 Aug 21 00:26:37 ip-10-188-2-114 nats-replicator[18223]: 2019/08/21 00:26:37.636792 [INF] error restarting streaming connection, will retry in 5000 milliseconds%!(EXTRA string=stan: clientID already registered)
Response from Ivan:
it means that the server still has this client ID and the client has the inbox running and can reply to the server check. It could be that the app got NATS disconnected and incorrectly tried to recreate the STAN connection (without closing the old one)?
Todo:
Look at logic when nats needs a restart, streaming that use that should be shutdown