Hello there!,
We have the performance problem described in the mail that I included.
Would you have any idea of what could be happening?
Any clue of what we should pursue on that system?

Why cannot we run the truss utility against the afsd processes?
We get a "truss: cannot control process 351" message.


                           thank you 


--------------------------------+---------------------------------------------
Ignacio Reguero: Unix Workstations Support
                 Computing and Networks Division
                 CERN,  European Laboratory for Particle Physics
--------------------------------+---------------------------------------------
Earth mail:                     |          E-mail:
   CERN / CN / DCI / USW        |      [EMAIL PROTECTED]
CH-1211 Geneva 23. Switzerland. |      [EMAIL PROTECTED]
Voice:                          |      [EMAIL PROTECTED]
   + 41 22 7674409              |  
Fax:                            |
   + 41 22 7678690              |
--------------------------------+---------------------------------------------


----- Begin Included Message -----

>From [EMAIL PROTECTED] Mon May  8 14:57:15 1995
X-Rerouted-To: [EMAIL PROTECTED] by the CERN Automatic Mail Router (v.2.0, July 
1993).
Date: Mon, 8 May 1995 14:57:07 +0200 (MET DST)
From: Alessandro Miotto <[EMAIL PROTECTED]>
To: Ignacio Reguero <[EMAIL PROTECTED]>
Subject: afsd behaviour on sun4m_53
Mime-Version: 1.0
Content-Type: TEXT/PLAIN; charset="US-ASCII"
Content-Length: 1496

I noticed a problem on two SPARCserver 20 model 612 running Solaris 2.3:
one of the afsd deamons starts eating a chunk of 20-30 CPU seconds 
every 10 minutes, blocking all other processes on the system during this 
period.
This behaviour is independent of any activity on AFS, and continues when 
the machines are idle.

On the same machines (whether related or not to the previous problem) RPC 
calls would time out because the associated deamons are not started by 
inetd.

[cms2] ~ > ps -ef | grep afsd
    root   351     1 80   Mar 13 ?          3:07 /usr/vice/etc/afsd 
-dcache 5000 -stat 5000 -volumes 128
    root   352     1 80   Mar 13 ?         3574:27 /usr/vice/etc/afsd 
-dcache 5000 -stat 5000 -volumes 128
                                           ^^^^^^^
    root   353     1 80   Mar 13 ?         12:04 /usr/vice/etc/afsd 
-dcache 5000 -stat 5000 -volumes 128
    root   354     1 80   Mar 13 ?         16:40 /usr/vice/etc/afsd 
-dcache 5000 -stat 5000 -volumes 128
    root   355     1 80   Mar 13 ?         45:20 /usr/vice/etc/afsd 
-dcache 5000 -stat 5000 -volumes 128
[cms2] ~ > uptime
  2:45pm  up 55 day(s), 18:39,  3 users,  load average: 0.08, 0.15, 0.20
             ^^^^^^^^^

---------------------------------------------------------------------------
 Alessandro Miotto - CN/DCI/UES      | Tel:    +41 22 767 9576
 CERN - European Laboratory          | E-mail: [EMAIL PROTECTED]
 for Particle Physics                |
 CH-1211 Geneve 23                   |



----- End Included Message -----


Reply via email to