[SERVER-24983] Remove method really slow Created: 11/Jul/16  Updated: 14/Jul/16  Resolved: 11/Jul/16

Status: Closed
Project: Core Server
Component/s: Admin, Querying
Affects Version/s: None
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: Ricardo Hilsenrath Assignee: Kelsey Schubert
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Duplicate
duplicates SERVER-24580 Improve performance when WiredTiger c... Closed
Operating System: ALL
Participants:

 Description   

My application was running fine since jun/27, and yesterday the remove() method became really slow

I'm using an i2.8xlarge in AWS

I tried to shutdown the service (db.stepDown() followed by db.shutdownServer()), and after 50 minutes it didn't shut down, I had to kill -9 <PID> to stop the process and reopen it, after restarting the service, the problem disappeared

I can provide the diagnotics folder and db.currentOp() output when the slowness was happening



 Comments   
Comment by Kelsey Schubert [ 11/Jul/16 ]

Hi ricardo_fanatee,

I've looked over the diagnostic.data you have provided. I see that the bytes currently in your WiredTiger slowly increased. At around 2016-07-10 23:58:00 the WiredTiger cache reached 95% utilization and application threads began evicting pages and your performance degraded.

MongoDB 3.2.8 contains a fix that significantly improves performance when the WiredTiger cache is full. Therefore, I would strongly recommend upgrading when it is released. Please see SERVER-24580 for additional details.

Please note that a release candidate, 3.2.8-rc1, for MongoDB 3.2.8 containing the fix for this issue is available for download now if you would like to see if it addresses your issue.

Kind regards,
Thomas

Comment by Ricardo Hilsenrath [ 11/Jul/16 ]

files uploaded

the problem happened approximately 15hours ago

Comment by Kelsey Schubert [ 11/Jul/16 ]

Hi ricardo_fanatee,

Thanks for reporting this behavior. I've created a secure upload portal for you to use to upload the files so we can continue to investigate. Would you please provide the logs as well so we can correlate the metrics in the diagnostic.data to the events recorded in the log?

Kind regards,
Thomas

Generated at Thu Feb 08 04:07:57 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.