Problem with RAM Usage

rObcq

• November 25, 2018 09:21

hi, all this problem started when godaddy made some changes (patches) and updates to the service, I have more than a year with its vps service with no problems. after they performed the changes, the site went down, fixed just repairing the db, then started tlooking at the ram ussage (free -m) and we had just 20 or 30 MB free used almost (total (2048MB). the thing is that we have this service for years now so when godaddy performed the patched they shudl change something, as you know godaddy has a very bad service so they are not able to help us since this is a VPS. so, what i found is that when i restart the server, the ram usages goes down drastically, after restarting we have 1500 1500 MB free and it starts climbing mb by mb so ussually after ONE DAY we are back having 20 or 30 free MBs. so I need to restart again, but this is not contsnatt, for example, yesterday i went to make backups and we had 670 free mbs, i started making the backups and it went down to 30 free MB wich ok, could be more or less normal, but then after ending the processes we dididnt get the MBs back, so i looks like it isa progressive loss of free mbs and the only way to solve this by now i restarting the server. Thank you very much for your help!

• 

  dalem

  • November 25, 2018 16:01

  Best to find out what process is using so much memory? What did they update? Note: it could be a bug in the kernel not releasing the memory perhaps upgrading or downgrading your kernel may help without more information no one here will be able to help

  0
• 

  rObcq

  • November 25, 2018 17:59

  thank you very much for your reply i try top -c and i see many processes but ussing just 0.1% or so, mysql is using the higerst % with just 3.5% also i have this procees, not sure if its normal or if it means something 978 mysql 20 0 1493m 71m 6080 S 2.0 3.5 1:50.54 /usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --plugin-dir=/usr/lib64/mysql/plugin --user=mysql --log-error=example.com godaddy changes: dont know they just informed about 3 patches (3 different days) they were performing. kernel 2.6.32-042stab133.2

  0
• 

  dalem

  • November 25, 2018 20:53

  I'm guessing they patched their node and not your server? Which should have had no effect on your VPS Your on openvz so kernel update does not apply free -m Adjust -n 10 as you see fit below to account for all the memory usage ps aux | awk '{print $2, $4, $11}' | sort -k2rn | head -n 10

  0
• 

  rObcq

  • November 25, 2018 22:05

  sns@server [~]# ps aux | awk '{print $2, $4, $11}' | sort -k2rn | head -n 10 1210 4.7 /usr/local/cpanel/3rdparty/perl/526/bin/perl 1212 4.7 spamd 1138 2.8 /usr/sbin/mysqld 2876 1.4 pkgacct 842 1.4 /usr/sbin/named 3222 1.3 pkgacct 1275 1.1 /usr/sbin/httpd 1362 1.1 cpsrvd 1404 0.7 tailwatchd 1280 0.5 /usr/local/cpanel/3rdparty/bin/perl sns@server [~]# free -m total used free shared buffers cached Mem: 2048 2048 0 54 0 1096 -/+ buffers/cache: 951 1096 Swap: 0 0 0

  0
• 

  dalem

  • November 26, 2018 01:31

  Looks like to me you have over 1G to spare

  0
• 

  cPanelLauren

  • November 27, 2018 16:50

  This usage doesn't look that abnormal does it get better/worse than this as time goes on? Historical statistics when it's at the worst might be useful. You can use sysstats for this and run something like: sar -r
  Prior to reboot

  0

Please sign in to leave a comment.