|
Servers Down and
then Up
Sounds like the stock
market eh? But again the stock market has been up more than down lately
:). The data servers crashed again sometime last last night. But they are
back up again.....but for how long??????
Screwy Stats
With the servers being up and down, it has been hit or miss with updating
the stats on this page. Even when the servers are up sometimes (like last
night) the stats pages dont update correctly. This all kind of sucks for
me because it makes my job that much harder (maybe they will increase my
pay? ;-). If the daily work units seem slow it is probably because of all
the server down time. I am not sure if they are finished with processing
all of the backlogs or not...I have a feeling they havent, we shall see in
the next few days.
Why Can't I
Connect? (and other
things)
Here are a couple of
snipets from SETI@Home staff member Eric Korpela I snipped from the alt.sci.seti
newsgroup a couple of days ago. After the server outage you may not have
been able to connect to the servers to send/receive data and here is one
reason why:
>but now I am home and
the same thing happened again. No connection possible. What's going
on????
For the bulk of yesterday we were maxed out on connections. We
were handing about 20 work units a second for most of the day, but were
getting 60 requests a second.
Eric
Also one of the things about the SETI
project is the possible lack of data to give to home members. Right now
there is a shortage because of the speed of the splitters which takes the
raw data and converts them to work units. But with the amount of users
maybe more data is needed:
>Why doesn't the SETI
team try to get permission to 'piggy-back' on other telescopes around
the world. It seems to me that the amount of data to process should be
less of a problem with the current support of 1.5 million people and
their computers.
We're working on it.....
Eric
Stats
SpeedRacer (aka. guru)
stretched his lead over the SouthPark Admin Crew a bit in the past couple
of days....but dont expect him to stretch it too much more. guru posted in
the Ars Technica Team thread:
Well that's it for seti.
Davycroc and I have started converting our seti machines over to RC5.
It's much easier to setup. It buffers the units, allows you to pull
units from multiple servers. It's almost to easy.
guru stated earlier in the
thread that they have over 400 machines running the SETI client. He has to
keep track of what machines are running what clients and has is set up
that two machines collected the resultes and upladed them. With the
upcoming change to the Version 2.0 client, he feels that this is the time
to make a switch. For him the RC5 client will be easier to setup and run
on his machines, and they seem to be having less problems and downtime
than the SETI project also.
Thanks for the ride(s)
guru, and good luck with the RC5 team!
On other top 20 moves: Guidant
is now in the #14 spot passing up DD2LM. Yesterday
we had a new member join the team and he immediately hopped into the #16
spot....I would like to welcome aboard Turbo
aka Flesh Rocket. Also FluffyPants
pulls one over the Dr.....Dr Xing that is ans slips by to #18. mjdewitt
and romkey
are in a battle to the death.....they have been within 10 work units for
about the past week....mjdewitt is holding romkey off though, even though
he only leads by 6 WUs.
Thank you and have
a nice day
Berkeley Servers
Back Up!
Things are getting back to
normal....The cgi-bin scripts seem to be working, as well as the stats
updates. I updated the top 100 results also...even though it may not be
that much of an update :-).
A 3 Hour Tour?
Apparently the server work
yesterday didnt go as smoothly as they wanted to and the data servers were
down for a lot longer than expected. Here is what they say:
December 16, 1999
On Tuesday, 12/15, we came down for what we had thought would be a 3
hour outage to rebuild the science database. The rebuild was necessary
because we had run out of database extents for the workunit table. The
Informix default extent size was much too low for us and we ran out
abruptly and without warning. After reloading the database from flat
files, the rebuilding of indices then took much longer than expected.
Finally, once the rebuild was complete, we suffered what appears to be a
malfunction in one of our RAID controllers, making it necessary to
restore the user root database from backup. Sun is currently looking at
the RAID problem.
The servers went back up
sometime around 4pm EST (I think?). But I have had a bit of problem trying
to upload the results. It appears to be a problem with everyone trying to
contact their servers at once. Luckily I still have a buffer of about 5
work units :-).
Stats Update
I updated the Club Team
Stats.....the Berkeley stats servers updated things ok late last night.
The numbers are kind of low mostly because the data servers were offline
almost all day yesterday and work units couldnt be processed. Individual
Team Lamb Chop stats were *not* updated. But that is not my fault.
Apparently the cgi-bin scripts are not working over there. I cannot access
the top 100 results for the team at the moment. The numbers that are on
the alternate page to get the numbers from have the exact same numbers
that I used for my update yesterday. Hopefully they will get this
straightened out soon.
Version 2.0
I received a copy (from Deimos)
of the email that the SETI guys will be sending out to everyone about the
Version 2.0 client:
--------------------------
PROJECT STATUS
With your assistance, SETI@home has been amazingly successful.
1,500,000people in 224 countries have downloaded the SETI@home
screensaver and together they have contributed 125,000 years of computer
time, with the common goal of finding the first sign of intelligent
extraterrestrial life.
We've been recording data at the Arecibo radio telescope since December
1998, and most of the data through May 1999 has now been analyzed.
100 million signals have been detected and stored in our database.
The second-phase processing, which rejects man-made interference and
looks for "repeat" signals, will start soon.
This is when we hope to actually detect an ET signal! Furthermore,
thanks to the abundance of computing power, we'll be augmenting
SETI@home to look for new types of signals. As the number of users has
grown, we've had to upgrade our servers several times to handle the
load.
Sometimes it's been hard to connect (sorry!) but our most recent
upgrades
have fixed this problem for now.
--------------------------
VERSION 2.0 COMING SOON
Most people are running version 1.06 of our software. Early next
year we'll release version 2.0, which has many new features and works
better through firewalls and proxies.
For more details, see
http://setiathome.berkeley.edu/version20.html
When version 2.0 is available for your type of computer, you'll be
notified
automatically the next time your computer completes a work unit.
When that
happens, you'll need to download the new version from our web site.
Dr. David P. Anderson
Project Director, SETI@home
http://setiathome.berkeley.edu
New No. 1!
SpeedRacer (aka
guru) posted a whopping 900+ work units today and took over the #1 spot
from the Southpark Admin Crew. I am about to watch the South Park Movie on
DVD tonight...maybe that will get them back into action! Congrats Speed!
Server Still Down
Looks like the Data Server over at the Berkeley offices is still down. Not
sure when it will be back up. Hopefully soon!
Server Down Time
From the Berkeley home
offices comes this news:
There will be a 3 hour
Data Server outage on 12/15/99 for database maintenance. This will run
from 9-12 PST.
SETI@Home Version
2.0
There was this news on the alt.sci.seti
newsgroup today:
VERSION 2.0 COMING SOON
Most people are running version 1.06 of our software.
Early next year we'll release version 2.0, which has many new features
and works better through firewalls and proxies.
For more details, see
http://setiathome.berkeley.edu/version20.html
When version 2.0 is available for your type of computer,
you'll be notified automatically the next time your computer
completes a work unit. When that happens, you'll need
to download the new version from our web site.
check out the news on their
site (with the link above). Unfortunately, the patch man Olli said that he
will come out with a patch for the new client also :-(.
Apparently this message is
going to be sent to every S@H member through the email address that they
used to sign up with. They are doing it alphabetically so if you are down
in the alphabet it may take a while :)
Top 20
There were a couple of
changes in the top 20 today. R.Marvin
streaks into the top 20....he did it with a flourish also. He jumped over
both Uncomm0n AND Grimas and is now at the #19 spot. With this, Grimas
drops out of the top 20. In another big move Vomesh
hopped past ojohnson AND OrlBamf and is now in the #11 spot. In other
news....romkey and
mjdewitt
are have been in such a battle that their lines in cannot be distinguished
from each other in the top 10 graph. mjdewitt though still holds on for
the #6 spot.
Finally...
I have finally gotten
around to updating all of the graphs. I changed around a couple of graphs.
The top teams graph is now a line graph. It doesnt show Art Bell because
they are way ahead of everyone else....but i have combined all of the
other top 13 teams into this graph. The other notable change is in the
"difference" graph. I reversed the y-axis on that graph, for
alittle bit more clarity. We will overtake either /. or MacAddict when the
line crosses the top of the graph instead of the bottom
(the way it was previously). I also updated the archived stats under the
club team stats heading. These graphs have a bit more data, and the graphs
themselves are a tad bit larger.
Weekly Stuff
SpeedRacer
ranked in again with the tops in work units done for the week. But for the
past couple of days he has slowed down a bit after taking the #2 spot.
Second in work units for the week is Vomesh.
He is followed by Gonzo
and C. Eric Smith.
The big mover for the week was svdsinner,
hopping over 20 spots this week. Congrats to them!
Milestones:
Over the past week we have had a bunch of new members to the 1000 work
unit club. They include acripps,
Trinity, Rat Bastard, Uncle Timbo, mzanthrope,
and The Weeblows.
New members to the top 100 include: SGWB,
Jonathan, and spagthorpe.
Congrats and welcome to the rarefied air!
There has also been a bit
of movement in the Top 20 in the past couple of days. romkey
and mjdewitt
are locked in a heated battle for the #7 spot overall. They have been
fighting back and forth. Today mjdewitt takes back the #7 spot, but that
may change tomorrow! Also, Vomesh
is still screaming up the charts.....in the past couple of days he has
moved past both D2TLM and Guidant and is now in the
#13 spot. Uncomm0n
and Grimas
are fighting it out for the #19 spot also. Uncomm0n is holding him
off....but maybe not for long. Both of them need to watch out because R.
Marvin is lurking in the #21 spot
and looking to move up!
Catching Up
I am still not sure how much more of a backlog the servers that the
Berkeley guys have left. So far Team Lamb Chop has had an estimated 11,000
WUs worth of backlog added to our team totals. I still think we have at
least that much more in backlogs yet to be posted to our team (maybe
more). So far since the backlog has been reduced Team Lamb Chop has gained
5500 WUs on MacAddict, and 3700 WU on /. I expect those numbers to at
least double before the entire backlog is flushed. We will have to wait
and see.
- News
Archives
|
