contrail-alarm-gen in a tight loop trying to connect to zk
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Juniper Openstack |
Fix Committed
|
High
|
Anish Mehta |
Bug Description
R3.0 Build 2713 Ubuntu 14.04 Kilo multi-node setup
Contrail alarm started reporting zk connection down on all the 3 controller nodes on this setup
root@katrina-
root 28264 0.0 0.0 11740 916 pts/0 R+ 00:34 0:00 grep --color=auto alarm
contrail 32407 92.9 0.6 317804 105928 ? Rl Feb15 534:14 /usr/bin/python /usr/bin/
root@katrina-
root@katrina-
173
root@katrina-
== Contrail Analytics ==
supervisor-
contrail-alarm-gen initializing (Zookeeper:
contrail-
contrail-
contrail-collector active
contrail-
contrail-
contrail-topology active
root@katrina-
2016-02-16 00:30:48,722 - WARN [NIOServerCxn.
2016-02-16 00:30:49,935 - WARN [NIOServerCxn.
2016-02-16 00:30:50,225 - WARN [NIOServerCxn.
2016-02-16 00:30:51,349 - WARN [NIOServerCxn.
2016-02-16 00:30:51,566 - WARN [NIOServerCxn.
2016-02-16 00:30:53,913 - WARN [NIOServerCxn.
2016-02-16 00:30:53,939 - WARN [NIOServerCxn.
2016-02-16 00:30:54,261 - WARN [NIOServerCxn.
2016-02-16 00:30:54,627 - WARN [NIOServerCxn.
2016-02-16 00:30:55,146 - WARN [NIOServerCxn.
root@katrina-
I'm also seeing this issue on my setup with the latest build 2713
@ nodeg13: contrail-alarm-gen initializing (Zookeeper: Zookeeper connection down)
{
"value":
[
{
"name": "nodea21",
"value":
{
"AlarmgenPa rtition" :
{
"inst_parts":
[
{
" instance" : "0",
" partitions" : [ ]
}
]
}
}
},
{
"name": "nodeg13",
"value":
{
"AlarmgenPa rtition" :
{
"inst_parts":
[
{
" instance" : "0",
" partitions" : [ ]
}
]
}
}
},
{
"name": "nodeg20",
"value":
{
"AlarmgenPa rtition" :
{
"inst_parts":
[
}
}
}
]
}