...finally became clear to me today, and I fixed the
problem.

I posted a few weeks ago about having results stuck in
SET Queue, which I use to manage my farm. I kept getting
"Connection reset" errors in the SQ log, but couldn't
figure out what was causing them--however, occasionally,
10-20 results would upload with no problem.

Finally, I checked the Results queue in SQ and noticed
that I had a bunch (about 23) of results from one animal
in the farm which were old--dating back to early November.
Opening the details on each of the results from that 
animal showed me that they all had an error (I think it
was 0xa0000064) in their status field. Every time SQ
tried to upload one of these results, the connection would
reset--the server was determining it was a bad result.

Apparently, this particular animal had indigestion of the
user_info.sah file; I ended up deleting all the stuck
results, and replaced the user_info.sah file with a copy
from a working system. Brought everything back up, and
results started uploading immediately, and now the Results
queue is empty.

While that was happening, I did some maintenance on the
log folders (copied them off into sub-folders by year) and
discovered over 7,000 copies of duplicate results plus another
1,600 or so of dropped results under the Client folder. Folders
that large are bad for performance, so cleaned those out and
did a defrag in Safe Mode; things are running smooth, now, so
I can work on those four animals that haven't been returning
results since the last power glitch.

I'd probably have found this faster if I hadn't been working
60+ hours/week through November and the first half of December.


Gary
61,749 results returned, 1631st place

-- 
Gary Heston  gheston@hiwaay.net 

"Sept. 11, 2001, already a day of immeasurable tragedy, cannot be
the day liberty perished in this country." Judge Gerald Tjoflat