[SERVER-26622] _recvChunkCommit should set errmsg when it encounters an error Created: 13/Oct/16 Updated: 05/Apr/17 Resolved: 29/Jan/17 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Diagnostics |
| Affects Version/s: | 3.4.0-rc0 |
| Fix Version/s: | 3.5.3 |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Randolph Tan | Assignee: | Nathan Myers |
| Resolution: | Done | Votes: | 0 |
| Labels: | None | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Backwards Compatibility: | Fully Compatible |
| Operating System: | ALL |
| Sprint: | Sharding 2017-01-02, Sharding 2017-02-13 |
| Participants: |
| Description |
|
Example of bad log message:
|
| Comments |
| Comment by Githook User [ 28/Jan/17 ] | |
|
Author: {u'username': u'nathan-myers-mongo', u'name': u'Nathan Myers', u'email': u'nathan.myers@10gen.com'}Message: | |
| Comment by Dianna Hohensee (Inactive) [ 12/Jan/17 ] | |
|
The UnknownError in Dan's comment looks like it was incidentally addressed by this commit. | |
| Comment by Randolph Tan [ 07/Nov/16 ] | |
|
kaloian.manassiev After reviewing the ticket again, I realize that I misread the code in getStatusFromCommandResult and conclude that there is nothing wrong with it. The problem is that _recvChunkCommit command returns ok: false without populating the errmsg field. This appears to be the combination of MigrationDestinationManager::startCommit not setting _errmsg when it returns false and MigrationDestinationManager::report selectively setting the errmsg field on the response. | |
| Comment by Kaloian Manassiev [ 28/Oct/16 ] | |
|
renctan, what kind of context could getStatusFromCommandResult add and from where would it take - can you elaborate in the ticket description? Does it make sense to change this ticket to be about auditing all sharding commands to ensure that we don't have code paths which do not return an error? | |
| Comment by Daniel Pasette (Inactive) [ 25/Oct/16 ] | |
|
Here's another one:
| |
| Comment by Mira Carey [ 24/Oct/16 ] | |
|
If there's some reason this should be on platforms, let us know. Otherwise I'll leave this ticket on the same backlog as the bf as it is a diagnostic change |