[SERVER-23614] High CPU Load on Idle/Prod Server Created: 07/Apr/16  Updated: 13/Apr/16  Resolved: 13/Apr/16

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

Type: Bug Priority: Major - P3
Reporter: Bodenhaltung [X] Assignee: Kelsey Schubert
Resolution: Duplicate Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Gentoo, Kernel 4.1.14.


Attachments: File diagnostic.data.tar.gz     File iostat.log.gz     File mongodb.log.gz    
Issue Links:
Duplicate
duplicates SERVER-2114 Don't use select timeouts for fast co... Closed
Operating System: ALL
Participants:

 Description   

Hi,

i have a fresh mongodb-3.2 running and a idle load from 1.x.

With a replset and live traffic the load increase to 4-5.

strace output:

select(8, [6 7], NULL, NULL, {0, 10000}) = 0 (Timeout)
select(8, [6 7], NULL, NULL, {0, 10000}) = 0 (Timeout)
select(8, [6 7], NULL, NULL, {0, 10000}) = 0 (Timeout)
select(8, [6 7], NULL, NULL, {0, 10000}) = 0 (Timeout)
select(8, [6 7], NULL, NULL, {0, 10000}) = 0 (Timeout)



 Comments   
Comment by Kelsey Schubert [ 13/Apr/16 ]

Hi Bodenhaltung,

Thank you for the clarification. From your description, I believe the issue that you are observing is tracked in SERVER-2114. Please feel free to vote for it and watch it for updates.

Kind regards,
Thomas

Comment by Bodenhaltung [X] [ 12/Apr/16 ]

Hi,

> Can you please confirm that you were experiencing this issue while you were collecting these stats?

Yes, i can confirm.

[...] can you please install MongoDB from Tarball and confirm the issue persists?

Yes, i can confirm, the issue is persists with the tarball mongod.

I can give you access on this machine, just send me your ssh key.

I have to correct me, the cpu load seems to be ok, but the load average is rising and with strace the mongod does not looking good.

Comment by Kelsey Schubert [ 12/Apr/16 ]

Hi Bodenhaltung,

I examined the iostat.log that you uploaded. It indicates that there was no CPU load while the data was being collected. Can you please confirm that you were experiencing this issue while you were collecting these stats?

Also, I noticed in your logs that you are using a debug build, can you please install MongoDB from Tarball and confirm the issue persists?

Thank you,
Thomas

Comment by Bodenhaltung [X] [ 09/Apr/16 ]

Hi,

After running this script for a day, please upload the following information:

  1. iostat.log
  2. diagnostic.data
  3. logs of the affected node

Ok, please see attached, i hope this will help.

Best Regards,

Comment by Kelsey Schubert [ 08/Apr/16 ]

Hi Bodenhaltung,

Thank you for reporting this issue. To get a better idea of what is going on here, can you please execute the following shell script:

delay=1
iostat -k -t -x ${delay:?} >iostat.log &

This will collect iostat data each second, and will help us to correlate the CPU numbers to events recorded in the diagnostic.data.

After running this script for a day, please upload the following information:

  1. iostat.log
  2. diagnostic.data
  3. logs of the affected node

Thank you,
Thomas

Generated at Thu Feb 08 04:03:55 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.