I have two nginx instances (nginx/1.0.15) on a 4GB RAM machine. Each
instance runs fewer than 25 requests as reported with stub_status on;
The problem is that once nginx is started from scratch it starts eating
(or reserving?) RAM progressively as reported by free -m, up to the
point where it leaves only 2-5mb left and then it doesnt go past that.
In about 7 days it has eating everything. Nginx doesnt crash, nor does
it touch swap but I definately feel it
compromises system resources to the point iam concerned as in a
production spike its wise to think the server has no room.
This is the exact same issue on 4 other nginx machines I have.
In order to start over again I must kill -QUIT the PIDs , delete the log
file (which is smaller then 800mb) and then all goes back to the cycle.
Please indicate me what pieces of information I can supply to debug this
issue.
Thanks
Joseph
Before anything I suggest you upgrade you nginx install… you’re
running
nginx/1.0.15, the last satble version is 1.2.8!
It might be some bug that was fixing in the mean time!
Hi
Are you sure it’s not the linux file/buffer cache that’s using all your
ram? (does ps/top show nginx or the worker processes using it directly?)
Linux and most/all other unix variants will fill up unused ram with
cached versions of the most recently used files so they don’t have to be
read from disk each time… it’s completely normal and expected
behaviour
Hello!,
This is the output of ps, iam having a hard time interpreting the memory
usage perhaps you can help me differ?
USER PID %CPU %MEM VSZ RSS TTY STAT START TIME COMMAND
root 18951 0.0 0.1 195892 6340 ? Ss Apr16 0:00 nginx:
master process /usr/sbin/nginx -c
/etc/nginx/modified/nginx-consolidated-5.conf
root 18977 0.0 0.1 197600 5188 ? Ss Apr16 0:00 nginx:
master process /usr/sbin/nginx -c
/etc/nginx/modified/nginx-consolidated-8.conf
nginx 18978 0.0 0.2 200096 9392 ? S Apr16 5:02 nginx:
worker process
nginx 18979 0.0 0.1 197604 5768 ? S Apr16 0:03 nginx:
cache manager process
nginx 23921 0.0 0.2 199892 9148 ? S Apr21 1:28 nginx:
worker process
nginx 23922 0.0 0.1 195896 4844 ? S Apr21 0:00 nginx:
cache manager process
root 24894 0.0 0.0 103244 832 pts/0 S+ 14:00 0:00 grep
nginx
[root]# free -m
total used free shared buffers
cached
Mem: 3631 1760 1871 0 307
1212
-/+ buffers/cache: 240 3391
Swap: 2047 0 2047
I have a cache explicitly set at the nginx.conf . My site is not that
big, the cache I have doesnt get pass 15mb.
Please let me know if this is suitable or if I should report
differently.
Thanks
Joseph
Date: Mon, 22 Apr 2013 20:32:31 +0100
From: [email protected]
To: [email protected]
Subject: Re: nginx eating all RAM, log files?
Hi
Are you sure it's not the linux file/buffer cache that's using all
your ram? (does ps/top show nginx or the worker processes using it
directly?)
Linux and most/all other unix variants will fill up unused ram with
cached versions of the most recently used files so they don't have
to be read from disk each time... it's completely normal and
expected behaviour :)
On 22/04/13 19:30, Joseph C.
wrote:
I have two nginx instances (nginx/1.0.15) on a 4GB
RAM machine. Each instance runs fewer than 25 requests as
reported with stub_status on;
The problem is that once nginx is started from scratch it starts
eating (or reserving?) RAM progressively as reported by free -m,
up to the point where it leaves only 2-5mb left and then it
doesnt go past that. In about 7 days it has eating everything.
Nginx doesnt crash, nor does it touch swap but I definately feel
it compromises system resources to the point iam concerned as in
a production spike its wise to think the server has no room.
This is the exact same issue on 4 other nginx machines I have.
In order to start over again I must kill -QUIT the PIDs , delete
the log file (which is smaller then 800mb) and then all goes
back to the cycle.
Please indicate me what pieces of information I can supply to
debug this issue.
Thanks
Joseph
_______________________________________________
nginx mailing list
[email protected]
http://mailman.nginx.org/mailman/listinfo/nginx
Maxim,
Thank you…!!
Regards,
Josephan/listinfo/nginx
man ps | grep RSS
man ps | grep VSZ
when using a tool like top: real_free = free + cached
you might want to try htop for a better continuous display
your ram is fine
regards,
mex
Posted at Nginx Forum:
Hello!
On Mon, Apr 22, 2013 at 06:30:41PM +0000, Joseph C. wrote:
a production spike its wise to think the server has no room.
This is the exact same issue on 4 other nginx machines I have.
In order to start over again I must kill -QUIT the PIDs , delete
the log file (which is smaller then 800mb) and then all goes
back to the cycle.
Please indicate me what pieces of information I can supply to
debug this issue.
As you are referring to “free -m” output, most likely this link
will help:
http://www.linuxatemyram.com/
In short: don’t panic, you RAM is fine.
–
Maxim D.
http://nginx.org/en/donation.html