[SERVER-19790] Provide mechanism to clear/acknowledge startup warnings Created: 06/Aug/15 Updated: 06/Dec/22 |
|
| Status: | Backlog |
| Project: | Core Server |
| Component/s: | Admin, Logging |
| Affects Version/s: | None |
| Fix Version/s: | None |
| Type: | New Feature | Priority: | Major - P3 |
| Reporter: | Victor Hooi | Assignee: | Backlog - Security Team |
| Resolution: | Unresolved | Votes: | 8 |
| Labels: | move-sec | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||||||||||||||
| Assigned Teams: |
Server Security
|
||||||||||||||||||||
| Participants: | |||||||||||||||||||||
| Case: | (copied to CRM) | ||||||||||||||||||||
| Description |
|
The mongod process can generate startup warnings under a number of scenarios (running as root, read-ahead values not optimal, THP enabled, _id index missing). These warnings appear both in the mongod.log logfile, as well as at the mongo shell prompt each time you login. (It's the second of these which this ticket concerns). However, these warnings are static - that is, they will not disappear even if the original condition is resolved, until the mongod process is restarted. Whilst some of these warnings refer to situations that would not change without a mongod restart anyway (e.g. read-ahead settings), others refer to situations that may be transient or that could have already been resolved (e.g. missing _id index, or THP settings). mongod should provide some mechanism to acknowledge/suppress or clear a warning for a condition that the user already knows has already been resolved, so that it would not appear on each subsequent mongo login. |
| Comments |
| Comment by Emilio Scalise [ 27/Sep/21 ] |
|
graham@brainnwave.com we are sorry to hear about your experiences with XFS. We recommend you reach out to your OS vendor for a deep RCA analysis and understand in detail what caused it (eg. storage issue, hardware issue and most unlikely XFS code bug). Generally XFS is considered more reliable and more robust than ext4. Please also note that ext4 has an hard limit in size (16TB). Beside your filesystem of choice, It's also very important that you build your MongoDB environments in a reliable and redundant way so storage / filesystem issues won't cause you a big impact and outages on your production environments. We can offer you a number of Services (eg. Support or Consulting Services) to guide you through this. Feel free to reach out to us using the Contact Us page. Kind regards, |
| Comment by Graham Jones [ 24/Sep/21 ] |
|
Upvoting on the basis that: 1) I have read that XFS is strongly recommend for WiredTiger. 2) I will never, ever, EVER use the XFS filesystem again. I've been burned by it and its catastrophic filesystem corruption that crippled a business for days whilst we restored and rebuilt petabytes of databases. No thanks, I'm happy with the increased performance risk of ext4 vs losing all my databases due to a power outage with XFS. So now please can it stop telling me? |
| Comment by Nic Cottrell [ 30/Jan/20 ] |
|
I'd like to see a new configuration option to disable specific startup warnings, so that remaining warnings are true positives and don't spam Ops Manager alerts etc. I envisage a new configuration file option like systemLog.hideWarnings which takes an array of strings like "dbPath-ext4" that disables just the ext4 dbPath volume warning, but leaves all others in place. This will require naming every warning and adding logic in startup_warnings_mongod.cpp and startup_warnings_common.cpp. |