I've tried everything I could find in the forum, wiki and google but still have the same problem. I've got 2 GB ram, after a few hours it is full and my ubuntu, server, no x crash.
What can I do now?
out of memory - Ubuntu 7.10 server - ZM 1.23.2
I'm watching the result of "top" (I think its the same like "htop") and can see how the memory is getting bigger and bigger .... until my whole memory 2GB.
It happens every time the cam is recording an event (function: Modect). Because I've placed the cam on the street = have 100 events per hour, after 2 or 3 houres the memory is full.
It happens every time the cam is recording an event (function: Modect). Because I've placed the cam on the street = have 100 events per hour, after 2 or 3 houres the memory is full.
I like htop because you can sort (you probably can with top, but I don't know how) and see what is taking up memory. Plus you get a nice graph at the top. But the bottom line is, you want to see what is using more and more memory in order to figure out where to start debugging (or searching these forums for more details).
Here's a little guide to the columns:
http://syscon.seartipy.com/use-htop-to- ... processes/
I guess mem% is the important one.
Here's a little guide to the columns:
http://syscon.seartipy.com/use-htop-to- ... processes/
I guess mem% is the important one.
I have installed htop. Please see the results on:
http://www.you2.ch/htop.png
Because you can not see on the picture:htop.png the whole line, here is the htop field Command of the marked line:
I've got many of those lines, so I think this is the problem? Can you help me to resolve my memory problem now?
http://www.you2.ch/htop.png
Because you can not see on the picture:htop.png the whole line, here is the htop field Command of the marked line:
Code: Select all
/usr/sbin/mysqld --basedir=/usr --datadir=/var/lib/mysql --user=mysql --pid-file=/var/run/mysqld/mysqld.pid --skip-external-locking --port=3306 --socket=/var/run/mysqld/mysqld.sockYou could try skip-innoDB in mysql:
uncomment "skip-innodb".
Restart mysql or your machine and see what happens.
Code: Select all
cd /etc/mysql
sudo nano my.cnf
Restart mysql or your machine and see what happens.
hmm... so in this way, what are the line ¦¦¦¦¦ before 222/2018MB - what is it showing?
And after starting the command: "top" I can read:
When I have top or htop open, after every motion detection I see ca. +10MB more of memory used and this is going until its full. After that I have a very slow system.
.... uncomment "skip-innodb". -> no changes.
I hope you've got another idea?
And after starting the command: "top" I can read:
Code: Select all
Mem: 2067268k total, 2027268k used, 58696 free.... uncomment "skip-innodb". -> no changes.
I hope you've got another idea?
I'm not a linux memory expert - all I know is, "it's complicated". Maybe top reports shared memory allocated for use? I just don't know. I'm guessing there are threads about this here, though - I've seen posts about memory usage problems before.
When I have a montage window open on a networked machine, I have just as many MySQL lines as you do (about 40 right now), though my "VIRT" is 53936 rather than 127M, which I think means that my mysql processes are only taking half the memory of yours. The important thing, I suspect, is to see what changes as your machine becomes slow. Does anything in the MEM% column go way up? You can sort on that column by pressing "M" (has to be capital M) when htop is running.
When I have a montage window open on a networked machine, I have just as many MySQL lines as you do (about 40 right now), though my "VIRT" is 53936 rather than 127M, which I think means that my mysql processes are only taking half the memory of yours. The important thing, I suspect, is to see what changes as your machine becomes slow. Does anything in the MEM% column go way up? You can sort on that column by pressing "M" (has to be capital M) when htop is running.
like I've been read in many threads and guides in this forum:
at the moment I have one cam: 420x288, b/w, PAL
at the end I will have one more with the same attributes.
Are they the right memory settings for this configuration?
Code: Select all
kernel.shmall = 134217728
kernel.shmmax = 134217728at the end I will have one more with the same attributes.
Are they the right memory settings for this configuration?
According to your picture above, htop says you aren't using any swap (0/1608MB). That's why I'd like to see a pic when things start to slow down; I think it might provide more information. My old ZM box with 512MB RAM always used *some* swap. Now with 1.5GB RAM I don't use any swap at all. The machine is noticeably faster.
Your shared memory settings look fine and are certainly adequate for 1 cam! I have 10 cams going with the same settings. Though I'll admit I've never completely understood the shared memory settings - there are some conflicting posts/wiki entries that seem to indicate one of those values should be divided by 4096 or something similar, but most posts just leave them the same.
Bottom line, on a headless Ubuntu box you shouldn't be losing memory every time there is an event. Do you have any filters running executables or something like that that might have a memory leak?
Your shared memory settings look fine and are certainly adequate for 1 cam! I have 10 cams going with the same settings. Though I'll admit I've never completely understood the shared memory settings - there are some conflicting posts/wiki entries that seem to indicate one of those values should be divided by 4096 or something similar, but most posts just leave them the same.
Bottom line, on a headless Ubuntu box you shouldn't be losing memory every time there is an event. Do you have any filters running executables or something like that that might have a memory leak?