inode leak in 2.2.14?

[Andrea Arcangeli]

On Mon, 31 Jan 2000, Oleg Drokin wrote:

>grow_inodes: inode-max limit reached
>After I did echo '16000' >/proc/sys/fs/inode-max
>thing worked again, I was curious and decided to look in /proc/sys/fs/inode-nr

Fine. Just set it up at boot to the amount of icache you want. 16000 is a
good value for a server (of course it depends how much memory do you have
:).

>I got "8150 27" from there, and there were no so much files open

No. Don't look at inode-nr. You must only verify if the "inode-max limit
reached" message triggers again or not with the new value.

The numbers in inode-nr have nothing to do with the number of inodes
actually in-use. It only say you how much memory is allocated for the
inode cache. It's more an information for developers.

>Server in question is not loaded squid+unfsd.

You probably needs lots of inodes for the squid files and for the network
connections, thus increasing inode-max when you get the message from the
kernel is the right thing to do and it's _not_ a sign of an inode leak.
It's instead the way we prevent inodes to leak ;).

Andrea


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo en vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/