From jzweizig@ligo.caltech.edu Wed May  4 19:09:11 2005
Date: Wed, 4 May 2005 16:09:09 -0700 (PDT)
From: John Zweizig <jzweizig@ligo.caltech.edu>
To: Peter Shawhan <shawhan_p@ligo.caltech.edu>
Cc: John Zweizig <zweizig_j@ligo.caltech.edu>, Keith Riles <kriles@umich.edu>
Subject: Re: H1UP_H2UNLOCKED and vice versa

Hi Peter,

Aha! There were two restarts of stone (running both lockloss processes) at
~793183380 and ~793426320 that effectively zeroed the lock state in the
trend files from the start of the hour to restart time. It should be
easy to replace these trends. Note that the first time span was before S4
started, so this was probably the "start a clean version" restart
before the run. I don't remember the why or wherefore but there seems to
have been a restart of all DMT processes on stone on Feb 25th at 20:i0PST.
I'll see what I can do to fix the trends and rerun the segment generation.
Ciao,

john

On Wed, 4 May 2005, Peter Shawhan wrote:

> Hi John,
> 
> When looking at DQ flags in v05, I found that there are some time intervals
> in the H1H2 data set which are flagged with H1UP_H2UNLOCKED and with
> H2UP_H1UNLOCKED.  It shouldn't be possible for either of these to be on
> during H1H2 running, let alone both.  Interestingly, they are both on
> for the same time intervals:
>     793180800-793183380
>     793425600-793426320
> Any ideas about how this could have happened?  I haven't yet checked for
> a problem with the base science-mode segment list, but the fact that BOTH
> are on at the same times seems to point to the flag generation, I think.
> 
> Peter
> 
> 
>