Some time ago I had this problem: Apache and logrotate configuration
In which my disk was full because of a huge error.log file. I solved the situation with a proper configuration of logrotate.
Now the problem occurred again (error.log of 50GB and 4 other compressed of almost 1GB, even though I set the maximum size of compressed files to 500MB.) So something went wrong with logrotate. Now I changed the configuration to a daily compression, set the number of compressed files to a maximum of 2 and with the same size. Hope it will work, I'm monitoring the situation (ah, running it manually it works properly).
So the first question is: what can cause logrotate not to work?
But the main question is: I realised Passenger is responsible for the huge log-files, I've been obligated to set the Passenger LogLevel to 0. It kept logging this each 5 or 10 milliseconds:
[ 2015-05-04 14:25:33.0003 21176/7f6938e91700 Pool2/Pool.h:790 ]: Analytics collection time...
[ 2015-05-04 14:25:33.0154 21176/7f6938e91700 Pool2/Pool.h:894 ]: Analytics collection done; next analytics collection in 4.985 sec
[ 2015-05-04 14:25:36.3909 21176/7f6938e50700 Pool2/Pool.h:675 ]: Garbage collection time...
[ 2015-05-04 14:25:36.3910 21176/7f6938e50700 Pool2/Pool.h:724 ]: Garbage collection done; next garbage collect in 46.617 sec
[ 2015-05-04 14:25:38.0003 21176/7f6938e91700 Pool2/Pool.h:790 ]: Analytics collection time...
[ 2015-05-04 14:25:38.0110 21176/7f6938e91700 Pool2/Pool.h:894 ]: Analytics collection done; next analytics collection in 4.989 sec
[ 2015-05-04 14:25:43.0003 21176/7f6938e91700 Pool2/Pool.h:790 ]: Analytics collection time...
[ 2015-05-04 14:25:43.0135 21176/7f6938e91700 Pool2/Pool.h:894 ]: Analytics collection done; next analytics collection in 4.986 sec
- What do these messages mean?
- And is it correct to let the loglevel to 0?
- Is it ok that even with LogLevel 0, it logs also the stdout console?
