cmd/jujud: machine agent keeps reconnecting to state

Bug #1169378 reported by Dave Cheney
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
juju-core
Won't Fix
High
Unassigned

Bug Description

Discovered during load testing, the machine agent keeps reconnecting to the state server

--

ubuntu@juju-goscale2-machine-27:~$ tail -f /var/log/juju/machine-27.log
2013/04/16 01:22:37 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:22:37 INFO state: connection established
2013/04/16 01:22:38 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:22:38 INFO state: connection established
2013/04/16 01:22:52 DEBUG state/presence: pinging "m#27" with seq=52
2013/04/16 01:23:08 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:23:08 INFO state: connection established
2013/04/16 01:23:08 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:23:08 INFO state: connection established
2013/04/16 01:23:15 DEBUG state/presence: pinging "m#27" with seq=52
2013/04/16 01:23:38 DEBUG state/presence: pinging "m#27" with seq=52
2013/04/16 01:23:38 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:23:38 INFO state: connection established
2013/04/16 01:23:39 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:23:39 INFO state: connection established
2013/04/16 01:24:01 DEBUG state/presence: pinging "m#27" with seq=52
2013/04/16 01:24:09 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:24:09 INFO state: connection established
2013/04/16 01:24:09 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:24:09 INFO state: connection established

Revision history for this message
Dave Cheney (dave-cheney) wrote :

Same with the unit agent.

--

ubuntu@juju-goscale2-machine-27:~$ tail -f /var/log/juju/unit-hadoop-slave-25.log
2013/04/16 01:24:01 DEBUG state/presence: pinging "u#hadoop-slave/25" with seq=53
2013/04/16 01:24:09 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:24:09 INFO state: connection established
2013/04/16 01:24:24 DEBUG state/presence: pinging "u#hadoop-slave/25" with seq=53
2013/04/16 01:24:40 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:24:40 INFO state: connection established
2013/04/16 01:24:47 DEBUG state/presence: pinging "u#hadoop-slave/25" with seq=53
2013/04/16 01:25:10 DEBUG state/presence: pinging "u#hadoop-slave/25" with seq=53
2013/04/16 01:25:10 INFO state: connecting to 15.185.169.145:37017
2013/04/16 01:25:10 INFO state: connection established

Revision history for this message
William Reade (fwereade) wrote :

This is mgo scanning for servers by reusing our supplied dial func, I think, I think. I've stopped it logging those attempts -- which is a concrete benefit, I think -- but I'm tentatively closing this WONTFIX until we can relate the reconnecting to another problem that justifies messing with it.

Changed in juju-core:
status: New → Won't Fix
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.