racluster exits

[Carter Bullard]

Hey Torbjorn,
I quietly uploaded argus-clients rc.68 last night that fixes a memory leak in racluster(), which may fix this issue.

Could you test this agianst your data?

   ftp://qosient.com/dev/argus-3.0/argus-clients-3.0.0.rc.68.tar.gz

The sleep(1) should be in the "retry the remote connection" code?  So that maybe a good amount of time to sleep?

Carter

 
Carter Bullard
QoSient LLC
150 E. 57th Street Suite 12D
New York, New York 10022
+1 212 588-9133 Phone
+1 212 588-9134 Fax

-----Original Message-----
From: Torbjorn.Wictorin at its.uu.se

Date: Tue, 15 Jan 2008 07:05:04 
To:Carter Bullard <carter at qosient.com>
Cc:Argus <argus-info at lists.andrew.cmu.edu>
Subject: Re: [ARGUS] racluster exits


Hi,

> I have made a number of changes to fix your racluster() problem,
> and did find a bug,  so we may have a reasonable solution.

there is still a problem. racluster does not die, but after some time
starts eating about 100% cpu and then does not produce any more output.

I 'touch'ed .devel and .debug and remaked in order to try to understand 
what happened. Then gdb says:

Starting program: 
/usr/local/src/argus/argus-clients-3.0.0.rc.67/bin/racluster -f 
/etc/racluster.conf -w /usr/local/argus/log/argus.log  -S 127.0.0.1:560
warning: Lowest section in system-supplied DSO at 0xffffe000 is .hash at 
ffffe0b4
[Thread debugging using libthread_db enabled]
Error while reading shared library symbols:
Cannot find new threads: generic error
Cannot find user-level thread for LWP 5859: generic error
(gdb)

I have not debugged threaded things in linux before, so it may be some 
very stupid mistake..

Btw, argus_client., line 1905:

================
       sleep(1);
    }
#endif
================

Should be nanosleep or some other threadsafe thing, I assume.


Torbjörn W