[Aide] Aide hanging - prelink block?

[John Horne]

Hello,

I currently have Aide 0.15 hanging on 3 CentOS 5.5 servers. The servers
use prelinking. Looking at one such server I see:

   ps auxww|grep aide
   root      2307  1.1  0.4 3109016 9404 pts/1    SN+  13:21   0:08 aide
      -u
   root     23397  0.0  0.0   4208   632 pts/2    SN+  13:34   0:00 grep
      aide

Running strace shows:

   strace -p 2307
   Process 2307 attached - interrupt to quit
   waitpid(16774,

Looking at the PID 16774 shows:

   ps p 16774
   PID TTY      STAT   TIME COMMAND
   16774 pts/1    SN+    0:00 /usr/sbin/prelink --verify /usr/bin/fexe

Running strace on that PID shows:

   strace -p 16774
   Process 16774 attached - interrupt to quit
   write(1, "/\232\0\0\0\0\0\0>\0\0\0\22\0\0\0V\5\0\0\0\0\0\0\214\0\0\0
      \21\0\0\0"..., 9588940

It seems that the prelink command is being blocked on a write to stdout.

Anyone any ideas about how this is occurring or what can be done about
it?

If a timeout on the prelink command (being used by Aide) was present
then that at least would prevent the systems from hanging up using Aide.
(We run aide regularly and when one gets into this state, all the
subsequent runs do too. After a weekend it means we have many processes,
on several systems, all just sitting there.)



Thanks,

John.

-- 
John Horne                   Tel: +44 (0)1752 587287
University of Plymouth, UK   Fax: +44 (0)1752 587001