-
Type: Bug
-
Resolution: Duplicate
-
Priority: Major - P3
-
None
-
Affects Version/s: 2.9.1
-
Component/s: Connection Management
-
None
When adding two mongo instances (2.4.2) to a replicate set containing two existing mongo instances (2.2.2) the following exception happened when adding to new instances to the replicate set:
Uncaught exception.com.mongodb.CommandResult$CommandFailure: command failed [count]:
{ "serverUsed" : "production4/172.16.25.200:27017" , "note" : "from execCommand" , "ok" : 0.0 , "errmsg" : "node is recovering"}at com.mongodb.CommandResult.getException(CommandResult.java:88) ~[mongo-java-driver-2.9.1.jar:na]
at com.mongodb.CommandResult.throwOnError(CommandResult.java:134) ~[mongo-java-driver-2.9.1.jar:na]
This seems to be the same issue as reported for the python driver here: https://jira.mongodb.org/browse/SERVER-7977
After this failure we first upgraded all the nodes to 2.4.2 and added the members to the replSet after the upgrade. This did work successfully. The bug https://jira.mongodb.org/browse/SERVER-7977 was reported to be actually a python driver bug but the fact that an upgrade to 2.4.2 solved the bug makes me wonder...
- duplicates
-
JAVA-753 Java Driver sending reads with NEAREST preference to nodes in RECOVERING state
- Closed
- related to
-
PYTHON-454 MongoReplicaSetClient should refresh immediately if we detect a secondary is recovering.
- Closed