[SERVER-27961] Help to analyze diagnostic.data after crash with OOM Created: 09/Feb/17  Updated: 27/Oct/23  Resolved: 09/Feb/17

Status: Closed
Project: Core Server
Component/s: None
Affects Version/s: 3.4.1
Fix Version/s: None

Type: Question Priority: Major - P3
Reporter: Grigorii Eremeev Assignee: Unassigned
Resolution: Works as Designed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Operating System: CentOS Linux 7 (Core)
CPE OS Name: cpe:/o:centos:centos:7
Kernel: Linux 3.14.31
Architecture: x86-64
-----------------------------------
WiredTiger
db version v3.4.1
git version: 5e103c4f5583e2566a45d740225dc250baacfbd7
OpenSSL version: OpenSSL 1.0.1e-fips 11 Feb 2013
allocator: tcmalloc
modules: none
build environment:
distmod: rhel70
distarch: x86_64
target_arch: x86_64


Attachments: File metrics.2017-02-09T10-34-34Z-00000     File mongod_shard.conf    
Participants:

 Description   

Hi!

Could you help me please to analyze diagnostic.data from shard?
Today my mongod process was killed by OOM. So far, I don't know what I should change that avoid this behavior. Maybe after you see diagnostic data and config for shards you will can advise me what I should change. Also, we don't use swap and we increase cacheSizeGB, available RAM is 47G. Thanks.



 Comments   
Comment by Alexander Gorrod [ 09/Feb/17 ]

Grigorii It appears as though you have configured MongoDB to have a 40GB WiredTiger cache, with 47GB of system memory. As a starting point that doesn't leave much memory available for other processes on the system, so I would recommend configuring a smaller cache size. From the data you uploaded, I can see that only 32GB of cache where in use when the process was terminated, so 40GB is significantly above the amount of available memory on your system.

It is also important to note that the WiredTiger cache size is not the total memory budget for MongoDB - it is only the memory allocated to cache data in order to minimize disk I/O.

I'd recommend configuring a smaller WiredTiger cache size, and monitoring your system memory usage to see whether the amount of memory used by all processes is approaching your available cache size.

This forum is for reporting bugs in MongoDB - wheras your question is about a configuration problem. If you have follow up questions you are better to ask on the MongoDB users google group.

Generated at Thu Feb 08 04:16:44 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.