sacretagent Posted June 27, 2013 Share Posted June 27, 2013 Since last week i get OOM messages on both of my servers now i don't want to point the finger yet to unraid that's why i ask if something was changed as last week i upgraded on both machines unraid plex and made changes to the cache dirs script .... which should normally result in less memory usage in my eyes i went from starting cached dirs with -m 5 -M 20 -d 5 -p 200 to -d 4 so less parameters and so i guess less memory used anyway i nearly never got OOM kills on any of these machines even with vbox installed ..... now i get like 3 a day most of the time Java gets killed from crashplan plugin sometimes python ... but i would like to know if you changed something at the memory parameters in the last release ? i am thinking going back to 12a to see if i still have them there .... modified to add syslog syslog-2013-06-27.zip Link to comment
limetech Posted June 27, 2013 Share Posted June 27, 2013 I made no changes regarding OOM. Why don't you set up a swap file? Link to comment
sacretagent Posted June 27, 2013 Author Share Posted June 27, 2013 I have one ... 8 gb big mm which reminds me i changed this too last week went from the unmenu swapfile to theone's plugin as it has a rc.d stop command can the issue be there ?? Link to comment
limetech Posted June 27, 2013 Share Posted June 27, 2013 I have one ... 8 gb big mm which reminds me i changed this too last week went from the unmenu swapfile to theone's plugin as it has a rc.d stop command can the issue be there ?? I'd say one of your plugins has a memory leak Link to comment
graywolf Posted June 27, 2013 Share Posted June 27, 2013 Since last week i get OOM messages on both of my servers now i don't want to point the finger yet to unraid that's why i ask if something was changed as last week i upgraded on both machines unraid plex and made changes to the cache dirs script .... which should normally result in less memory usage in my eyes i went from starting cached dirs with -m 5 -M 20 -d 5 -p 200 to -d 4 so less parameters and so i guess less memory used anyway i nearly never got OOM kills on any of these machines even with vbox installed ..... now i get like 3 a day most of the time Java gets killed from crashplan plugin sometimes python ... but i would like to know if you changed something at the memory parameters in the last release ? i am thinking going back to 12a to see if i still have them there .... modified to add syslog Have you reverted your cachedir parms back to previous settings? If so, do you still get the OOM? If not, reset your cachedir parms back, restart your array (so you have a clean shutdown/startup) and see if you still get the OOM BTW, less parms does NOT mean less memory, all depends upon what the default values of the parms you no longer specified compared to what you had them set. I don't run cachedirs so not familiar with the various parms and their interactions Link to comment
Recommended Posts
Archived
This topic is now archived and is closed to further replies.