Subject: Re: 2.5-meter Obs Log (MJD 51574)
Date: Mon, 31 Jan 2000 14:54:57 -0600
From: Ronald Rechenmacher
>
> SDSSHOST crashed at 18:30. It rebooted by itself, but then we couldn't do an
> initNight...id1 wouldn't boot. We had to replace the id1-0 drive and then
> initNight worked. We are not sure what caused the problem, but here are
> things that happened just before 'HOST crashed:
> 	1) Both simulator FOXIs were removed.
> 	2) The spectro DA crate was then rebooted
> Could these have somehow caused the crash? Can we start up astroDA without
> the simulator FOXIs?
In the SYSLOG, there are ptvme problems basically 10 minutes before
the reboot/crash AND a couple of problems for 20 minutes after the crash.
The astroda log does not show any relevant events after the log and starts
at the reboot (the previous log info is probably lost).  If the spectro
crate was power cycled, this may have contributed to the cause. A crate
connected to the PTLINK should not be power cycled while the ptvme link
is active. The simulator board will not know/care if FOXI boards are attached
or not.
I have talked to Jim Annis about storing astroda logs (trying not to delete
any log data) for about a week.
Jim,
I tried to submit to the gnat but got:
   Error
   Your problem report has not been sent.
   Fix the following problems, then submit the problem report again:
       Submitter-Id is 'unknown'
Should someone make me a real person (as far as the sdss gnats system is
concerned)?
Craig and I just talked further about the ptlink problem and concluded
it was either a) a crate power up glitch or b) the spectro power
cycle caused the spectro pt card to break.  He will be doing some
ptvmeTest tests and report.
--Ron