bin/115814: /etc/rc.d/named restart fails sometimes

[João Carlos Mendes Luís]

>Number:         115814
>Category:       bin
>Synopsis:       /etc/rc.d/named restart fails sometimes
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    freebsd-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Sat Aug 25 15:50:01 GMT 2007
>Closed-Date:
>Last-Modified:
>Originator:     João Carlos Mendes Luís
>Release:        FreeBSD 6.2-STABLE i386
>Organization:
Barradev Tecnologia Ltda.
>Environment:
System: FreeBSD 6.2-STABLE Wed Jun 13 14:32:29 BRT 2007, i386

chrooted named
master for a few domains, slave for a few others, recursive for a small network

>Description:


# /etc/rc.d/named restart
Stopping named.
named already running? (pid=48587).
# /etc/rc.d/named restart
Stopping named: rndc failed, trying killall: No matching processes were found
Starting named.
# /etc/rc.d/named restart
Stopping named.
Starting named.
# /etc/rc.d/named restart
Stopping named.
Starting named.

>How-To-Repeat:

As seen above, just after restart the problem goes away.

Probably a delay when dying and many info still in memory.

>Fix:

I am no expert in rc-subr, but the fix is to make sure named has died before exiting the stop subroutine.

I usually do some kill -0 to check if a pid is still alive.

IMHO, every program that is stopped with kill(2) should have that kind of check.
Maybe this could be a new routine to rc.subr.

>Release-Note:
>Audit-Trail:
>Unformatted: