[SERVER-44539] Re-enable "missing RWC" logging Created: 11/Nov/19  Updated: 06/Dec/22

Status: Backlog
Project: Core Server
Component/s: Diagnostics, Sharding
Affects Version/s: None
Fix Version/s: None

Type: Improvement Priority: Major - P3
Reporter: Kevin Pulo Assignee: Backlog - Replication Team
Resolution: Unresolved Votes: 0
Labels: PM-900-Fallout, sharding-wfbf-day
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Depends
depends on SERVER-45692 Ensure all internal inter-node comman... Closed
Related
related to SERVER-55699 Add an explicit write concern to inte... Closed
is related to SERVER-44887 Complete TODO listed in SERVER-43712 Closed
Assigned Teams:
Replication
Sprint: Sharding 2020-01-27, Sharding 2020-02-10
Participants:

 Description   

After both SERVER-43712 and SERVER-43126 have been completed, re-enable the "missing RWC" logging (on shard/config servers) once we're sure that they won't result in logspam (or in a way which avoids that). This may also require adjusting any jstests that issue commands directly against shard/config servers.



 Comments   
Comment by Kaloian Manassiev [ 14/Jan/20 ]

Also I wanted to add that in the presence of the testing done under SERVER-45098, enabling this logging is a little bit less pressing, because at least we have a confirmation that the commands are apply the defaults correctly. This logging was intended to ensure that any new commands we might add won't forget to add the defaults handling.

Comment by Kevin Pulo [ 14/Jan/20 ]

Yes, the idea is to re-enable the logging, and then fix up any excessively verbose logging that results. The expected sources are:

  1. User operations that are inadvertently not sending RWC to the shards. Some of these are in the process of being fixed on SERVER-45098. You should also be aware of SERVER-45549.
  2. Internal operations on mongos/shards which don't specify RWC (like the config.cache.chunks.XXX initialisation you found). These should have an appropriate RWC explicitly added to them (most likely {w:1} and {level:"local"}).
  3. jstests that do operations directly on shards. This might include ShardingTest setup/initialisation/shutdown. These should have an appropriate RWC explicitly added to them (almost certainly {w:1} and {level:"local"}).
Comment by Jack Mulrow [ 13/Jan/20 ]

kevin.pulo, kaloian.manassiev, I ran an evergreen patch with the logging enabled, and it looks like we still have a good number of internal operations that don't specify a read/write concern and trigger the logging (e.g. creating an index on a config.cache.chunks.XXX collection).

Just to confirm before I spend more time on this - was this ticket meant to include finding and fixing up these operations? Thanks!

Comment by Kaloian Manassiev [ 02/Dec/19 ]

Now that both the dependent tickets have been closed, assigning this to you Kev to evaluate whether we should re-enable that logging.

Generated at Thu Feb 08 05:06:15 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.