|
It's Update Time!
First up it is the long awaited return of the weekly
stats! Yay! There is a double dose of the weekly stats seeing
that I totally forgot to do them last week. You can check out this
week's weekly stats here, and the
previous week's weeklies here.
Caching Updates
There is a new beta for Seti Queue out now...It is now up to the Beta4e
iteration and can be downloaded from the Seti
Queue pages. now with the revamping of the entire Seti Queue
program, it has made the Seti Queue guide that was listed on the tips
section here outdated. Former TLC member Geordie
has written a new
guide for installation and setup of the improved Seti Queue.
It is a good read if you want to set it up and find out about all the new
and nifty features!
Keeping in the lines of new addon programs, there is a new
version of Seti Monitor
out. Please make note of the new URL for Seti Monitor!
Some New Menus
Been messing round with the menus, and there is a new look for them.
They are drop down menus which allows you to navigate through the stats
from nearly any page. They render slightly different in Netscape and
IE, but the functionality is the same. Let me know if there are any
problems with them. Hopefully I will have some more news sometime
later tonight :)
One of the things you may notice that is added is a
category for Overall
Charts. These charts are some stats on the top 200 overall
teams in the S@H project. There are different categories for the
stats and each category is sortable. You can view the pages either
by choosing it from the "Overall Charts" drop down menu on the
left, or just by clicking on the column heading (the heading is underlined
and can be clicked to take you to that page). One stat on the pages
to make note of in the next couple of weeks is the Overtake column.
Right now TLC is set to pass Sun Microsystems for the #1 spot overall in
less than a month! Kick Ass!
The Young Step-Sister
I received an email from SabreWulf pointing
out one of the caching programs that was missing from the links section of
the site here. It is the Step-Sister of caching programs,
SetiGate. It doesn't get the "press" that either SETI
Driver or SetiQueue does, but nonetheless is a worthy option for your
caching needs. I will let SabreWulf tell you a bit about it:
For the past week I have been trying a
WU caching program, called SetiGate.
I don't know if either of you has come across it. The url to the
site is http://www.gss.myweb.nl/seti/
and the link to the download
is http://members.telocity.com/hatster/setigate/download.htm.
I know that it might not be my place to promote some else's
package. But from what I seen so far, it seem rock solid.
The only problems I had with it have been user related (RTFM) stuff. I
will do all the regular stuff that we got used to hiding and starting
the cmd line dos box (local machine only), on the remotes they still
need to be run manually (I think)...
It will act as a proxy client / portal
for your seti, you must have a static IP address for the server
PC, which for most home/work users should be quite easy to
set-up. The work situation might be a bit more tricky, they're
going to have to sweet talk their network admin department. It
also helps if you only have a single connection to the internet as the
unit are in the one place and the results are return to same
machine. Then it's a simple matter of clicking on connect and
the send and receive. Sending and downloading the work units at
once. Your even luckier if you have a permanent connect
because it will do this with out you having to a thing at all.
In my case I was running Setidriver
with various amounts of cache on numerous machine in the office.
Which you might appreciate, could be a pain in the arse. Never quite
knowing whether they're running or not. But with SetiGate this
doesn't seem to be a problem. It will monitor every client that it
set-up to feed units to.
One thing I nearly forget to
mention was that it will feed multiple users so you could have one cache
for everyone at work. I haven't seen any limits on the number of
users and clients that it can monitor.
I have tried various other client
server programs, this one seem the easiest to set-up.
Sent Work Units....How
They Do That?
There had been some stuff going round alt.sci.seti about people wondering
how many times they send out work units and how long the work units stay
around until they are deleted from their servers. Eric
Korpela responded with some information (to straighten things
up) on how they deal with the handing out of work units, and how many
results they get returned for each work unit:
>But now on to the final
question: The S@h homepage states "We send out
>each work unit multiple times in order to make sure that the data is
>processed correctly". How often is multiple?
It depends. We mark workunits for deletion from the send queue
after the second result is received. The actual deletion happens
in chunks of 100,000 units (assuming that many are ready to be deleted,
fewer if fewer are ready) when the splitters have gotten to the point of
nearly filling the disks. So it's possible for a workunit to done only
twice, and it's possible for it to be done 5 times if the demand for
workunits is exceptionally high or if it doesn't take very long to
compute because of RFI. Right now we average about 3.3 result per
workunit (based upon workunits #81000000 to #82000000).
Eric
CGI Page Downtime and
the Future
About a week ago there were problems with the download of the stats from
the cgi pages on the Berkeley servers. There may have been a problem
with the databases, but that seems to be straightened out at the
moment. There also may have been another reason why the cgi pages
were down....and it may be explained by a post Matt
Lebofsky had on alt.sci.seti a couple of days ago:
As a team we are focused on
cleaning up the science database regime for the most part. The next big
milestone will be when we have a clean database for swift data analysis
as well as an on-line science database for quick queries (and, yes,
that's when we'll turn the last 10 workunit pages back on, as well as
many other web goodies).
So...take what you want from that...the cgi pages may have
been quite slow because the science database was pretty clogged, or they
may have taken the cgi pages offline for just routine maintenance which
had nothing to do with the cgi pages being slow. One of the
"web goodies" seems to have shown up already. They have
redesigned the front page of the site, but the other pages on the site
seem to be the same. I'm not sure if I like the new front page...it
seems a bit cluttered for my liking. The question that Matt
responded to was actually about the possibility of a 3.04 version...but
Matt stated that talks of a 3.04 version hasn't come up in their meetings
yet.
Partial Stats For Today
The download of the stats yesterday took significantly longer (over 2x the
time) than normal. That was the main reason for the stats being up
late yesterday. Today's stats are:
Why is this? Yesterday's download of the stats from
the Berkeley cgi pages were quite slow. Usually this means a problem
with their cgi server and/or their stats/science database(s). I
emailed Eric Korpela, letting him know that there may be something up with
their servers (if they didn't already know), to maybe nip a problem in the
bud. He replied that they would take a look. Today when I was
pulling the stats, I noticed that they weren't doing anything too fast,
and then found out the problem. The cgi pages over on the S@H site
are down. They are down now (~10pm Eastern) and were down when I
started to pull the stats at 4pm Eastern also. I can only assume
that there *is* a problem with their server/database and they are working
on it.
This means that I could not pull the stats from the top
200 teams....including TLC. But I had to have *something* up...so I
did the stats tonight from the normal TLC .html page on their
servers. The problem with this is that the .html pages aren't
updated real time, and also depends on the cgi/database also.....so even
though i pulled the data somewhere around 9:00 PM, the data from those
pages were from before 4pm eastern....so they aren't quite accurate for
the time that is listed on the page.
Stats, Scripting,
Incompatibilities, Oh My!
If you have read the March 9 news you will know that I was working on a
way to automate the download and upload of the stats for the site
here. I DO have something working, and it worked well for Monday,
when the S@H servers weren't having any problems. I was working
towards some more "enhancements" for the pages also, and that
lead me to trying to incorporate cascading style sheets in the
pages. First off I would like to say, with the way I do stats and
save the .html pages in Excel it is not a trivial thing to set up css for
the stats pages. I found out a way to do it through Front Page, AND
IT WORKED! Sorta. I quickly was notified that the stats pages
were crashing Netscape machines, and if it didn't do that, there was some
major hard drive thrashing going on. I do have to admit that I don't
know exactly how the css pages work, but I did quickly figure out what was
causing the problems.
Using css(s) significantly made the stats pages smaller,
each chart page was reduced from a whopping 180+kb down to a relatively
svelte 80kb. The way I had the css pages set up, several different
chart pages were linked to a single css page. Well that one css page
was around 300kb in size. I had fired up system monitor and
tried looking at a chart page using IE5.5 and also Netscape 4.75.
Using IE5.5, I had no problems loading, and the available RAM drop was
quite small. On the other hand, while using Netscape, it tried to
read the css sheet, and while trying to load the 300kb CSS, the available
RAM dropped like a rock. Were talking about a drop of 50-60MB.
This was probably why the Netscape machines were having problems.
Now don't even ask how a 300kb file can lay waste to 50MB of RAM.
Needless to say, I changed the way I was doing the pages back to the way
before. I may have a way to alleviate the css problem, but that is
going to have to wait till this weekend at least.
WARNING!!!!
I am going to be out of town tomorrow, and tomorrow's stats update
(Saturday the 10th) will be an attempt at a totally automated stats
download, calculation, and upload. I am not sure if it is going to
crash in the middle or anything, so there is a possibility that it may not
work. That would be no stats for tomorrow :/. I am going to
test this in a couple of minutes, so the stats in about an hour or so may
look a bit funny. One thing to note about the automatic upload is
that I will not be able to use the MS Office HTML filter, to filter out
some of the MS code bloat, so the files may be a bit larger and take a bit
longer to download (especially the member charts). Hopefully all
will go well and the stats will be posted normal tomorrow :) UPDATE!!!
It looks like the test run went ok, with one small error that wont
effect the automation for tomorrow :). Sorry for the site having an
"intermediate" update...Things should be ok at the next stats
update later today (Saturday).
Milestone? No a 3
Millstone!
Team Lamb Chop has done it again....this time with *clean* results.
Sometime yesterday TLC passed the 3 million work
unit mark! WOOHOO! Taking a look at the team numbers, TLC
should be passing another milestone within a week....that would be the 4000
year mark in CPU time contributed to the project. Cool
Beans. Each TLC member should turn to their brothers and sisters and
arms and congratulate each other for being the most kick ass team out
there :)
More Wallpaper Linkage
While on IRC tonight, ColinT pointed
me to some more wallpaper quality pics for your PC. Check out the
pics from National Space Science Data Center Mars
Photo Gallery. The first three photos on the page are a
mosaic of several different views of Mars. If you scroll down a bit
more you will see some pictures of the famous "face" in
Mars. Make sure you also check out the view of the "face"
taken from the Mars Global Surveyor....hrmmmmmm doesn't seem like it looks
like a "face" anymore eh? Or is it just a big
government cover-up! ;)
You may want to check out some of their other photo
galleries to find some things of interest for you also. I
haven't checked the entire site out yet...but will do so soon.
More Stuff
In addition ColinT pointed me out to
this "hands on" sort of distributed project. It is the NASA
Ames Clickworkers study. The study uses you (if you chose to
accept the mission) to help mark craters on the Mars surface, and also to
classify the different types of craters that have been found. One
thing tha is cool about the study is that you can take a look at many
different high resolution pictures of Mars taken by the Mars Global
Surveyor. After looking at some of the pictures, I have come to the
conclusion that life must have existed on Mars at one time, in one
form or another. One word of warning if you want to click, click,
click away...it can be taxing on your eyes and your clicking
fingers...make sure you take regular breaks! Cratering isn't just
for Quake3 Arena anymore!
Random Stats Stuff...
I was taking a look at the numbers that the top 200 teams have put up
tonight...and noticed a couple of random things. TLC has contributed
the second highest amount of CPU time for any team. Art Bell is
tops, in that regard (by a large margin). Of course that number that
AB has put up is due to their over 13,000 members, compared to TLC's ~4000
members. AB has an average time per work unit of nearly 23
hours/WU...compared to TLC's average of 11.75. Hypothetically, if AB
had the average time of TLC, they would currently have over 4 million work
units processed and would be sitting at #1 right now.
In the top 200 teams, there are 48 other teams who have
average work unit times faster than TLC. Why is that? i think
it can be explained in two ways 1) about 10 or so of those teams are of
the corporate type, with powerful machines....Sun has over 3 million work
units, but only a 7.5 hour average. and #2) is the GUI effect.
When I first started keeping stats I believe the average tie for TLC was
up around the 20.5 hour/WU mark...TLC had quite a bit of work units
amassed at that time and mostly through the GUI client. The average
WU time rapidly decreased since then and bottomed out around the 11 hour
mark (before the 3.03 client showed up)...at that time our daily averages
were in the 7-8 hour range!. The majority of teams with better
time/WU averages above TLC right now are teams who came onto the scene
pretty late and rode the tide of faster processors and faster clients to
fast averages.
So what does that all mean? Nothing I guess....just
writing something for the page. Sorry for rambling, it is only
3:30am :)
Link 'O the Day
I was watching the Discovery Channel last night, and they had a show about
the Hubble Telescope. I had seen parts of this show before, but
watching it this time and the last time lead me to scrounge up some pics
from the telescope. I went back to my favorites folder and clicked
on the shortcut for Hubble
Space Telescope Public Pictures. I know that there are other
sites which may look snazzier, but they all link back to these pages where
you can download the pictures.
What can I say? Some of the pictures from the HST
are just absolutely beautiful. Below are a few of my favorites from
the site:
Ghostly
Reflections in the Pleiades
Close-up
View of a Reflection Nebula in Orion
Embryonic
Stars Emerge from Interstellar ``EGGs''
Super-Sharp
View Of The Doomed Star Eta Carinae
Light
and Shadow in the Carina Nebula
The last link above is a picture that I decided to make
the background picture on my computer here. Many of the pictures on
the HST site have different formats and different sized pictures that you
can download. Some of the pics are pretty large so you can easily
convert them to wallpaper pics for your machine. Shown below on the
left is the full pic (albeit a bit smaller than on the site) of my current
wallpaper. Right before I started writing this, I noticed my
favorite part of the picture. That is shown on the bottom
right.
Yes we have the FIRST confirmed signal from
extraterrestrial beings. Yes the cheaters in the S@H project are
getting the finger from ET, several million light years away.
Cruncher o' The Week!
If you haven't noticed, the S@H guys picks out an active user every week
to be the "Cruncher of the Week". This week's cruncher
of the week turns out to be TLC's own Diesel71!
Congrats! and crunch on!
Lingering Problems
It seems that things are nearly back to normal after the down time for the
Berkeley network, but I did run into some dropped connections when trying
to upload today. They don't really know why there are dropped
connections but here is what they have to say about it:
We came back on line on Saturday,
March 3rd, and rather quickly recovered from the back log of clients
trying to send results and get new workunits. We are operating normally
now, though we are currently dropping a few connections a second, and
are trying to determine why (possibly a disk mounting issue).
While talking about connections...Ever wonder how much
connections and bandwidth the S@H servers handle through a day? I
found a couple of links on the alt.sci.seti newsgroup over the weekend
which will let you see how much data they push on a regular basis.
You can check out graphs showing the data they push and the amounts of
packets they push on this
page. The graphs of interest are the second set of three
down (fastethernet1_1_0: 169.229.1.49: ssl-evans p2p fe link). When
the S@H servers went back online, their servers were pegging their
bandwidth limit of about 30 Million bits/sec. It has settled down to
about 20-25 Million bits/sec now.
If you want to see how bad their bandwidth situation was
pre 3.03 release? Take a look at this page
of graphs. The bandwidth cap for the project is 30M
bits/sec. From March to the end of November their bandwidth used
steadily increased. In the first part of December, the bandwidth
they were pushing was pegging their bandwidth limit. At least you
can see the problems that they were facing, and the reason why they deemed
that the increased science in version 3.03 was necessary to help alleviate
their bandwidth problems.
New Science Newsletter
Last week they snuck one in on us....the S@H guys posted Science
Newsletter #6. The newsletter outlines some of the ways that
they use to help distinguish possible signals produced by ET from random
noise and terrestrial radio interference. Go take a look if you are
so inclined.
SetiQueue
For the past month or so Ken Reneris
has been working on a new version of SetiQueue.
So far the new version is still in beta, but it does work :). It is
in the 4th revision and if you don't mind using beta software for caching,
I suggest that you give it a look. I have been using it since the
second or so revision, and it has quite a bit of features over the old
client. You can choose from a GUI version, and also a command line
version for installation. There are way too many changes,
improvements to go over here so I suggest you take a look at the version
history and also the page detailing some of the features.
I forgot to add....there is a good discussion going on
about the new version of SetiQ over at the Ars Distributed Forum. It
is a good read, and maybe you can add some input on what you would like
the new SetiQ to look like. Check it out here.
Stats R Up
I pulled the stats somewhere around 10:00 PM (Eastern Time) tonight...I
wanted to get some stats up sometime tonight :). I am sure that most
people haven't been able to upload their cached work units
yet....hopefully tomorrow's update will be more robust!
WOO...They're Back!
Or at least for a while I guess....I was just able to access
the S@H site, even though it is a bit tougher trying to upload results
right now :/. Here is the message that is on the S@H page right now:
BERKELEY NETWORK OUTAGE NEWS:
Around 11:00 GMT (3:00am PST) on Tuesday, February 27, 2001, network
fibers were broken, cutting off the entire Space Sciences Laboratory and
Lawrence Berkeley Labs from the internet. The SETI@home website and data
server were unaccessible for several days during the entire length of
the outage. Due to the large bandwidth SETI@home requires, it may be as
long as 24-48 hours after the Space Lab comes back on-line before the
data server is fully functional and accepting all connections.
Be patient, hopefully things will be back to normal soon.
Is Today the Day?
Somewhere on the West Coast, forces are being mobilized to affect repairs
on the offending busted fiber optic cables. The site is still not up
yet (3pm Eastern Time), but I wouldn't expect the site to be up until
later in the day anyways. Earlier today there was a post sort of
giving more insight at the looks of the site that they are trying to
reach...and I will pass it along to you now:
Susan wrote:
> Interesting
situation. I've been to both the Berkeley campus and the new
> Space Science Lab up on the hill and although I remember a rather
windy road
> up to it I thought the whole area was pretty much built-up and
accessible.
You probably went up Centennial Road. It runs through Strawberry
Canyon, and the area right around it is pretty developed. But just
off the road it gets pretty wooded and there are ravines and canyons
(and mountain lions have been spotted off the road near the Botanical
Garden). The location of this fiber is in a wooded area near the
"Big C." Friday morning, an LBL staffer drove a truck over to
the area to prepare for the arrival of the contractor, and the truck
started sliding sideways down the hill into a ditch. So the
contractor brought a 4x4 out...and also slid into the ditch.
> Also, why can't the
repair crew walk a block to the site--I suppose they
> need electrical power that the conduit itself cannot supply?
As someone else mentioned, there is some large equipment needed at the
site.
> And now, on second
thought, I'm really confused. Did 'they' pull two new
> cables that now must be spliced where the original damage occurred
or are they
> only replacing half the cable to the damaged area--and if so--why
is only half
> being replaced?
It took me a while to get the story straight. The fiber was
damaged in a manhole vault, so they pulled new bundles through the
conduit between the manhole directly above and below the manhole where
the break occurred. Now they need to splice it at those two
manholes.
> Will the crew work
the problem over the weekend in the rain once they
> find figure out how to get to it?
I hope so. The weather cleared up in the afternoon and the radio
said it's not going to start raining until the afternoon (and Mike
Peckner is never wrong).
I need to go into campus Saturday to do some work, so maybe I'll try to
stop up there and see how it's going.
michael
When things get back going on the S@H servers I may wait a
bit till I pull stats for the site. I haven't decided on what I will
do yet...but check in later tonight to see if they are up :)
A Comedy of Errors?
OK....you would think that having internet (and whatever other)
connections to an important building at a prestigious university would get
you priority in repairs. Then on top of that your project is
dependent on connections from 2 million plus people, don't you think that
problems would get fixed ASAP no matter what the barriers? I guess
the Cal. Berkeley grounds crew can't negotiate around some rain dampened
terrain. As a result, the S@H servers will
not be back online till tomorrow (Saturday) at the earliest.
Here is the scoop.
The fiber pull has been completed
and the final step is to splice the new fiber back into the main LBL-UCB
bundle.
Unfortunately...
Due to the rains in the Bay Area yesterday and this morning, when the
contractor tried to access the site where the splice needs to be made,
their trucks (including 4x4s) became stuck. They have been unable
to reach the location.
It gets wackier: tomorrow, they're going to rent ATVs so they can get to
the site. Of course, that means we won't be back until tomorrow.
Bah! Date Pushed
Back.
Ok this is the latest news....looks like things are back to
Friday.....LATE Friday till things will be back up:
This morning we (CNS) were told
that the repairs would be finished sometime today. Unfortunately that's
not the case. The carrier has finished pulling new cable, but
now LBNL has to splice the new cable into their own fiber plant. This
process will not be completed until late in the day Friday, 2 March.
Oh well..
While The Servers Are
Down...
I think it is time to catch up on some space related news. There has
been quite a bit of interesting things going on in the past couple of
weeks....and here are some of them.
-
Space Exploration --
BushWhacked: In the new budget going to be set forth by
Bush axes several projects. Gone is the Pluto
Mission for now, also a Solar Probe mission. There
appears to be increased funding for robotic exploration of Mars
though. Also the X-33
Spaceplane got the axe. This was a project to develop
a single stage plane that could make it to orbit and back.
There also seems the possibility that some of the modules of the International
Space Station may get scaled back, or even axed
completely. I think that many of these cuts are pretty
sad. For way too long the US has seemed to lose its
foresight. I sit at home watching Babylon 5 almost every night
thinking about exploration to the outer regions of space, but the
stars seem to keep getting farther and farther out of our
reach. Unfortunately, with the axing of these projects we also
loose technology discovered in the process, and other things that
may be passed down to the consumer level. Who knows, maybe one
day we shall venture outside of Earth's gravity.
-
Planetary Society Helps
Launch Solar Sail: The Planetary Society along with
their sponsor Cosmos Studios are nearing the launch of a Solar
Sail. This sail will be 30 meters in diameter.
It will be launched from a submarine, in a converted ICBM. The
launch is scheduled to take place around April 19, and the mission
is planned to start around October-December.
-
Trigger For Mass
Extinction: Scientists think they have found the
trigger for the largest
extinction on Earth. Mind you this is not the one
surrounding the extinction of the dinosaurs. The evidence
seems to be found within buckyballs. Read the link above for
more info.
-
Life on Mars?: Woah....you
may have heard about the possibility of life on Mars a year or go,
but those experiment's conclusions were put in doubt by many
others. There seems to be better evidence that life had
existed on Mars from more recent
experiments..
-
Recipe for Life Elsewhere?:
First there was an experiment that showed that cells could
possibly be formed from material found in deep space. Next
there was an announcement that scientists found water vapor
and carbon molecules surrounding young and dying stars. Then
you can add in some findings that suggest that donut
shaped dust clouds around young stars are evidence that many
stars should have Earth sized planets. What do you get?
Ok not EVIDENCE that there is life in other solar systems, but you
do have increasing amounts of evidence that there may be (or
should?) be life elsewhere in the universe. This should be a
very exciting time for space exploration, and increasing our
knowledge of things outside the Earth. Unfortunately, with
budget cuts the excitement is tempered by a government who seems
unwilling to look outside our tiny little planet.
Date Pushed Up.
The message now on the S@H site shows that the expected date for the fixes
being pushed up to today. The message now reads:
Contractors are pulling new cable. We now
expect that service to SSL will be restored sometime today, Thursday, 1
March 2001.
The servers aren't up yet (5:00pm EST)...and of course we're not sure
when they will finally be up. I still have a couple of hours till I
am hurting....I have a small handful of work units left in my SetiQ stash
:). Lets hope it will be soon. Then the mad rush for
downloading work units will begin!
Servers Down Until
Friday???
It looks like there was a message on the Berkeley communications and
networks site, which now appears when you try to access the main S@H
site. Here it goes:
Fiber cut silences
SETI@Home
At about 3:30 AM PST on 27 February an
optical fiber cable connecting the U.C. Berkeley campus with the
Lawrence Berkeley National Laboratory was cut, apparently by vandals
trying to "salvage" copper from other nearby cables.
The broken fiber carries data and
voice connections for LBNL and also for the Space Sciences Lab. SSL is
where the SETI@Home project is located, so the millions of participants
helping to analyze data have been unable to contact the SETI@Home
servers for more than a day.
Contractors are pulling new cable now.
It's expected that service to SSL will be restored by Friday, 2 March
2001. We'll update this page as we learn more about the progress of the
repairs.
Ug. Friday eh? Of course, when I installed the
new SetiQueue beta, I decided to down the amount of work units I needed to
cache seeing that the S@H servers were down for several hours at
most. I only had a 3 day stock of work units in my queue...I
am going to run out of work units sometime tomorrow night. Oh
well. It has been a bit over a year since a day has gone by that I
had not crunched a work unit on my machines here, and that streak may come
to an end.
I guess Michael below was correct seeing he said "DO
NOT QUOTE!!!". It wasn't a construction crew that caused the
damage....but more like a "DESTRUCTION" crew. It is kind
of sad that the entire project would have to be shut down because of this.
-Front Page |
