2.0.6: very hard overload after upgrade from 2.0.5a

[Lack Mr G M]

Eugeny Kuzakov <core at coredumped.null.ru> wrote (and I precised):

> smbstat shows this:
>    (11 connected shares by 8 systems).
> 
> "ps ax | grep smbd | grep -v grep" shows this
>  2589  ??  R    3108:46.48 smbd
> 11889  ??  R     13:12.33 smbd
> 12072  ??  R     12:39.44 smbd
> ...
> 24414  ??  I      0:07.12 smbd
> 26314  ??  S      0:01.20 smbd
>
> What can I do?

   If you compare the pids you will see that those which smbstat knows
nothing about have been using CP time - hours of it (months for the one
at the top!).  Possibly disappeared clients causing a problem.

   You could try adding this to the [global] section of your config:

      socket options = SO_KEEPALIVE

--
--------- Gordon Lack --------------- gml4410 at ggr.co.uk  ------------
This message *may* reflect my personal opinion.  It is *not* intended
to reflect those of my employer, or anyone else.