-
Type:
Question
-
Resolution: Done
-
Priority:
Major - P3
-
None
-
Affects Version/s: 3.0.1
-
Component/s: Replication, Sharding
-
None
Hello.
What's mean in mongos log
2015-04-01T16:15:52.695+0200 W SHARDING [Balancer] multiVersion cluster detected, my version is 3.0.1
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] attr0001 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] events0001 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] events0002 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] events0003 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] goods0000 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] killmail0001 is at version 3.0.1
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] photo0001 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] photo0002 is at version 2.6.9
2015-04-01T16:15:52.695+0200 I SHARDING [Balancer] webscan0001 is at version 3.0.1
2015-04-01T16:01:23.803+0200 W NETWORK [conn828122] Primary for photo0001/db0012.xxxxx.com:xxxx,db0013.xxxx.com:xxxx,db0014.xxxx.com:xxxx was down before, bypassing setShardVersion. The
local replica set view and targeting may be stale.iew and targeting may be stale.
for all rs nodes once in 5 minutes ffter upgrading 2.6.8 to 3.0.1 ?
In one rs I see in log connections from another rs, such as:
2015-03-28T22:47:02.033+0100 [conn28] scoped connection to db0012.xxxx.com:xxxx not being returned to the pool 2015-03-28T22:47:03.448+0100 [conn32] scoped connection to db0012.xxxx.com:xxxx not being returned to the pool 2015-03-28T22:47:03.769+0100 [conn53] scoped connection to db0012.xxxx.com:xxxx not being returned to the pool 2015-03-28T22:47:04.094+0100 [conn4461] scoped connection to db0014.xxxx.com:xxxx not being returned to the pool 2015-03-28T22:47:04.212+0100 [conn4461] scoped connection to db0012.xxxx.com:xxxx not being returned to the pool 2015-03-28T22:47:04.792+0100 [conn1489] scoped connection to db0014.xxxx.com:xxxx not being returned to the pool
Looks like not OK.