Hey,
I have a cluster of 4 nodes, that has been running fine for months now. And, all of a sudden when I am trying to remove a node from the replicaset, the primary is giving me the following assertion. I am running mongo 1.8.1 on all the nodes on ubuntu 10.04.
(primary_mongo_shell) >>rs.remove('host01')
{
"assertion" : "bad repl set config?",
"assertionCode" : 13122,
"errmsg" : "db assertion failure",
"ok" : 0
}
(primary_mongo_shell) >>rs.conf()
{
"_id" : "mongo",
"version" : 5000,
"members" : [
{
"_id" : 18,
"host" : "host05:27017"
},
{
"_id" : 27,
"host" : "host04"
},
{
"_id" : 31,
"host" : "host06"
},
{
"_id" : 33,
"host" : "host01"
}
]
}
(primary_mongo_log):
Thu Aug 16 10:16:15 [conn1460978] query admin.$cmd ntoreturn:1 assertion command: { replSetReconfig: { _id: "cellar", version: 5001, members: [ { host: "host05:27017", _id: 18 }, { host: "host04", _id: 27 }, { host: "host06", _id: 31 }, { host: "host01", _id: 33 } ] } } reslen:141 1557ms
also, host05 is the primary.
Any ideas? How can I get out of this situation?
Thanks,
--Eytan