[syslog-ng] Leaking objects in size-4096 cache in slabinfo

Doug Warner silfreed at silfreed.net
Wed Dec 16 15:54:10 CET 2009


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 12/15/2009 03:17 PM, Balazs Scheidler wrote:
> Possibly a socket structure (but AFAIR it has its own slab). Can you
> check if:
>   * syslog-ng has too many open sockets (visible for example with lsof)

We have about 290 open TCP connections.

>   * whether it reads its input queue (although network packets would
> again not show up in the slab), using netstat -antp

Most have smallish Recv-Q's (<2000KB), but a couple hosts have receive queues
of 50KB, and one of >1MB.  Is this an indication that my syslog-ng instance is
overloaded and can't get the data out to disk fast enough?

>   * possibly check the network statistics: netstat -ns

Nothing looks too odd here.

Thanks for your help debugging this issue!

- -Doug
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iD8DBQFLKPSSJV36su0A0xIRApNsAKDcdXer/qYeXLsiGXj3fHnnb3PrswCgmrV+
Txho9qhyIJxvXVeoirEgSws=
=75bV
-----END PGP SIGNATURE-----


More information about the syslog-ng mailing list