Fixed one bug, got some more for ya though :)

[Carter Bullard]

Hey Chris,
Hmmm, this is a problem where the argus server side process
that was sending records to your remote ra() had either its
queue get toooo long, or the maximum number of non-fatal
errors on the output socket was reached.  Argus has to decide
at some point that the remote cannot keep up, or is not there,
and so when a threshold is reached, Argus has to let the
remote socket go.

There may be a bug in this, so let me a take a look.

One thing you can do is to increase the values of the thresholds.
If you double the value for ARGUS_MAXERROR in ArgusUtil.c,
that will help a bit.  Also raising the watermark for printing
out when queues are getting big will help, so increase
the value of ARGUS_LISTREPORTLEN to something like 50,000.
This is also in ArgusUtil.c.
   

Carter

Carter Bullard
QoSient, LLC
300 E. 56th Street, Suite 18K
New York, New York  10022

carter at qosient.com
Phone +1 212 588-9133
Fax   +1 212 588-9134
http://qosient.com

-----Original Message-----
From: Chris Newton [mailto:newton at unb.ca] 
Sent: Thursday, June 14, 2001 7:26 AM
To: Carter Bullard
Subject: Fixed one bug, got some more for ya though :)


Hi Carter.  Seems that recent set of fixes you gave me fixed ra from
dieing 
very frequently (dozens of times per day).  It ran flawlessly for 36
hours 
without one hiccup... 'cept this ;)

  This morning I came in, and noticed that ra was no longer dumping log 
files... running 'top' showed ra not using any CPU.  So, I went over to
the 
argus server, and checked it... argus was over there running... but, not

gathering any data, it seemed.  Checking /var/log/messages on the argus
server 
found:

Jun 14 06:48:56 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
10001
Jun 14 06:49:26 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
11455
Jun 14 06:49:56 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
12919
Jun 14 06:50:27 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
14371
Jun 14 06:50:57 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
15813
Jun 14 06:51:27 epic argus[13037]: ArgusWriteOutSocket(0x81755a0) Queue
Count 
17271
Jun 14 06:51:51 epic argus[13037]: ArgusHandleClientData:
ArgusWriteOutSocket 
failed Resource temporarily unavailable


now.. the weird thing... I didnt restart the argus server to fix the
problem.  
I restarted ra.  Killed ra, started a new one, things are back to
normal.


I'm wondering too..  I noticed that argus frequently (on my network,
which is 
violent as hell... great for testing), argus spits out messages about
Queue 
Count getting high...  I'm wondering what packet levels are required to
cause 
this?  And, if argus is tuned to send flows out the 'out door' fast
enough?  
I'm wondering this, because... I might soon have the opportunity to test
this 
setup on our telco's link.. which, might be running around 150 Mbit
(running 
at, not connected at :))  Ideas?

Thanks Carter,

Chris

_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/

Chris Newton, Systems Analyst
Computing Services, University of New Brunswick
newton at unb.ca 506-447-3212(voice) 506-453-3590(fax)

"The best way to have a good idea is to have a lot of ideas." Linus
Pauling (1901 - 1994) US chemist