[JAVA-531] Driver allows connection to multiple replica sets without any errors or warnings Created: 07/Mar/12 Updated: 29/Aug/12 Resolved: 24/Jul/12 |
|
| Status: | Closed |
| Project: | Java Driver |
| Component/s: | Cluster Management |
| Affects Version/s: | 2.6.5 |
| Fix Version/s: | 2.9.0 |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Mark Miller | Assignee: | Steve Briskin (Inactive) |
| Resolution: | Done | Votes: | 0 |
| Labels: | replicaset | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Environment: |
Centos on AWS |
||
| Description |
|
The driver allows a connection string with a number of hosts, but the hosts can belong to multiple replica sets. The situation occurred when we used two nodes to create another replica set, but failed to change the connection string. No error's were reported by the driver, instead strange application behaviour was experienced as sometimes nodes 5 and 6 were queried, but contained no relevant data. I know there is a replicaSet option, which I believe causing more checking, but I would think that the default behaviour would be to check that all hosts were in the same replica sets, and to only be able to see one primary. |
| Comments |
| Comment by Jeffrey Yemin [ 29/Aug/12 ] |
|
Closing as part of 2.9.0 release process. |
| Comment by auto [ 24/Jul/12 ] |
|
Author: {u'date': u'2012-07-24T08:21:18-07:00', u'name': u'Steve Briskin', u'email': u'steve.briskin@10gen.com'}Message: |
| Comment by auto [ 24/Jul/12 ] |
|
Author: {u'date': u'2012-07-23T11:52:09-07:00', u'email': u'steve.briskin@10gen.com', u'name': u'Steve Briskin'}Message: |
| Comment by Mark Miller [ 07/Mar/12 ] |
|
mongod -version is db version v2.0.0, pdfile version 4.5 |
| Comment by Scott Hernandez (Inactive) [ 07/Mar/12 ] |
|
What server version are you using? |
| Comment by Jeffrey Yemin [ 07/Mar/12 ] |
|
Do you see any SEVERE logging lines containing the string "mis match set name old:"? That would show that the driver is at least detecting the problem. But looking at the code, it looks like it won't stop the driver from preceding despite detecting the problem. We should fix this. |