Re: process table

[Andrew Findlay <andrew.findlay@skills-1st.co.uk>]

On Tue, Aug 06, 2002 at 07:54:52AM -0400, Andrew Bacchi wrote:
> 
> I happened to restart my ldap service this morning and noticed a lot of 
> dead processes that were hanging around.  Is this a known openldap 
> problem, or is the system to blame?  Is this in any way slowing down the 
> system?  Is restarting on a regular basis the recommended fix?  
> 
> Here is a small sample of the restart.  System is RH Linux 7.3, Openldap 
> 2.0.23.  Thanks
> 
> /etc/init.d/ldap: kill: (29844) - No such process
> /etc/init.d/ldap: kill: (29843) - No such process
> /etc/init.d/ldap: kill: (29842) - No such process
> /etc/init.d/ldap: kill: (29841) - No such process
> /etc/init.d/ldap: kill: (29838) - No such process
> /etc/init.d/ldap: kill: (29835) - No such process
> /etc/init.d/ldap: kill: (29834) - No such process

/etc/init.d/ldap on RedHat uses 'killproc /usr/sbin/slapd' to stop the
server. This will try to kill each PID that has a matching process
name. SLAPD is multi-threaded, and Linux allocates one PID per active
thread so killproc tries to kill them all. The error messages you see
are because the first signal kills all threads, so the remaining ones
have nothing to kill.

It would be better if /etc/init.d/ldap caused SLAPD to write a PID
file at startup, and then used that at shutdown time to decide which
process to kill.

Andrew
-- 
-----------------------------------------------------------------------
|                 From Andrew Findlay, Skills 1st Ltd                 |
| Consultant in large-scale systems, networks, and directory services |
|        Andrew.Findlay@skills-1st.co.uk       +44 1628 782565        |
-----------------------------------------------------------------------