[MXS-3567] Potential memory leak Created: 2021-05-24  Updated: 2021-09-12  Resolved: 2021-09-01

Status: Closed
Project: MariaDB MaxScale
Component/s: maxctrl
Affects Version/s: 2.5.10
Fix Version/s: 2.5.12

Type: Bug Priority: Major
Reporter: Chin Yuen Assignee: markus makela
Resolution: Fixed Votes: 0
Labels: None
Environment:

Red Hat Enterprise Linux 8.3 (Ootpa)
Intel(R) Xeon(R) Gold 6278C CPU @ 2.60GHz
8gb ram

Attachments: PNG File maxscle.png     JPEG File os memory.jpg     JPEG File os_.jpg    
Sprint: MXS-SPRINT-139

 Description   

memory usage seems to grow overtime in maxscale,
2 os memory graph one is taken around 11 am another one around 12 pm
growing from 41% to 54% in m one hour time.

while the backend galera cluster stable at 5.5% ram usage.

 Comments   
Comment by markus makela [ 2021-07-12 ]

Can you attach the MaxScale configuration with all passwords and other sensitive information removed?

Comment by markus makela [ 2021-09-01 ]

cyngate is it possible for you to let us know which modules in MaxScale you are using? This might be specific to some module so being able to reduce the search space would be helpful.

Comment by Chin Yuen [ 2021-09-01 ]

Hi Markus, Sorry for late reply, we tried to patch to version 2.5.12 and it fixed the issue after the maxscale support team principle debug the issue and found out and told us that the EXPLAIN leak memory MXS-3566 was the root cause

thanks and regards
Ng

Comment by markus makela [ 2021-09-01 ]

OK, that's great to hear. I'll close this issue as fixed in 2.5.12 then.

Generated at Thu Feb 08 04:22:21 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.