HP3000-L Archives

February 1995, Week 2

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Jeff Woods <[log in to unmask]>
Reply To:
Date:
Thu, 9 Feb 1995 14:16:47 CST
Content-Type:
text/plain
Parts/Attachments:
text/plain (49 lines)
Pete Crosby <[log in to unmask]> wrote:
|Ken Sletten <[log in to unmask]> wrote:
[snip: text about repeated unsuccessful powerfail recovery]
|
|            The number one cause of unrecovered PFAILS is sloppy
|power when power is restored. After a power outage all kinds of
|things are powering up and causing undetectable brownouts and
|sags and surges. A good line conditioner (isolation AND regulation)
|will prevent these anomalies from reaching the system and its
|core components.
|
|            The next time you encounter a PFAIL and the system
|seems to hang up, try generating a PFAIL of your own. Pull the
|power cord on the CPU (to kick it into battery back-up), cycle
|power on the console and external discs and tapes, then plug the
|CPU back in. This will cause the system to go through PFAIL
|RECOVERY with, hopefully, clean power and in most cases will
|bring the system back to life without the need for a restart.
|
|            There is always the possibility that there is a CPU
|or power supply problem with your system and that probably needs
|to be checked out. However, I have used the above technique, very
|successfully, for a lot of years on a lot of different 3000's and
|PA3000's. I hope this helps.
 
Advice from the "For What It's Worth Department":
 
During my years as a 3000 SysAdmin, whenever I had a powerfail that lasted
long enough for me to get to the hardware (which was far too often), I would
immediately pull the power cord on the CPU(s) then turn off the power switch
on all the external peripherals.  After the power came back on (and stayed on
long enough for me to believe that it was staying on), I would power on all
the peripherals and after they were spun up, back online, and such then plug
in the CPU power cord again.  Not only did this prevent bouncing the system
power like a yo-yo during XMgr recovery (which caused us to have to re-INSTALL
an MPE/XL system at least once), but I also believe it was less stress on the
hardware.
 
This seems a bit extreme, but the behavior was developed while working for
nine years in an office where a main power feed for the utility company was
at the end of the block and next to an access road of a highway being widened.
For a few years, the power failures were almost weekly and lasted anywhere
from less than one second to hours.
--
[log in to unmask] (Jeff Woods at Unison Software) [PGP key available]
We have a criminal jury system which is superior to any in the world; and its
efficiency is only marred by the difficulty of finding twelve men every day
who don't know anything and can't read.  -- Mark Twain (1835-1910)

ATOM RSS1 RSS2