HP OpenVMS Systems

SECURITY_SERVER and pagefaulting?

» close window

The Question is:

 
Why is there an excessive page fault rate on the SECURITY_SERVER process?
I am newly responsible for an OpenVMS cluster, my first experience with the
Alpha platform. We are running OpenVMS 7.1 on an Alpha 2100 and Vax 4300 -
separate system disks. The Alpha is far and away the heavy hitter. The Vax
is used primarily for custome
r service logins (few). The cluster was rebooted two days ago and the
Alpha's process page fault rate is already in excess of 4.5 million with
only 4400 i/os (via SHOW SYSTEM). SHOW/PROCESS/CONTINOUS shows the process
consistantly in HIB state, changing c
ur/base priority states, doing only 19 i/o s and racking up page faults. The
Vax's process looks "normal." NCP on the Alpha says "DECnet for OpenVMS AXP
V6.2." On the Vax it reads DECnet for OpenVMS VAX 6.1."  There are both
NET$PROXY (shown being accesse
d by the server via SHOW DEVICE/FILES) and NETPROXY on a device these two
boxes share. The files only contain three or four records but are different.
Suggestions would be appreciated. Thanks.
 

The Answer is :

 
  This pagefaulting behavior appears rather anomolous.
 
  Please contact the customer support center for assistance.
 
  Expect to be asked for information on the number of proxies on the system,
  on the contents of the intrusion database, and on the amount of proxy- and
  intrusion-related activity.  You may also be asked about your system
  parameter settings particularly around memory reclaimation, and around
  your use of MODPARAMS.DAT and AUTOGEN.
 

  

     

     
answer written or last revised on ( 29-SEP-1998 )
     
» close window