Join BAUT teams for BOINC(Einstein-,SETI-,LHC-,...)

[azazul]

Congratulations to SarahMc, our newest member to get 100,000 credits in E@H.

[Ken Vogt]

Also belated congrats to The Rock and Sarah.

----
I'm sure it's noted somewhere on their site, but the µfluids@home project is currently at the alpha stage. I apologize for not knowing this when I posted about it.

As we all have learned with BOINC, a project being alpha means far less than it would for an ordinary Windows program, I have very rarely if ever seen the BOINC core client crash: if a project crashes, BOINC just switches to another one, or downloads a new WU. And I have definitely never seen a BOINC error so bad as to bring down Windows.

Nevertheless there may be issues with running µfluids - mainly some WUs erroring out. For more details, I've started a thread about it here on azazul's board.

[azazul]

I didn't realize this at the time that I mentioned it either. So far, I have had a few problems with not getting credit and some wu's creating errors. But, there has been no computer problems, BOINC has taken care of the errors on its own.

[Wolverine]

Since February, I've only returned one client error running E@H, and I'm pretty sure that was due to an unrelated crash on my machine.

[gopher65]

I haven't had any problems that weren't related to me yanking out my HD and putting a new one in without letting BOINC finish its work unit.

EDIT: Seperate issue here, but due to the nature of these boards it is impossible to pick up an old thread and read through it (well, impossible for me). Could we maybe get a message saying "join BAUT teams for Boinc...etc" sticked and LOCKED at the top of one of these forums? Or maybe the BA and the UT guy could put a permanant link to the teams and azazul's forum on their main pages.

[Ken Vogt]

Sorry for the delayed reply...

I agree with both Wolverine's and gopher's experience: other than the occasional foolish move by me, or the bad disk drives I had, BOINC has run almost without ever crashing. Considering it has had by far the most CPU cycles, that is something.

EDIT: Seperate issue here, but due to the nature of these boards it is impossible to pick up an old thread and read through it (well, impossible for me). Could we maybe get a message saying "join BAUT teams for Boinc...etc" sticked and LOCKED at the top of one of these forums? Or maybe the BA and the UT guy could put a permanant link to the teams and azazul's forum on their main pages.

It is of course up to azazul to decide if he wants to ask for these, but I think this thread (especially thanks to Wolverine for its stickiness) works pretty well. The first post is still pretty current, and directs new people to the right places, and the thread is short enough to be skimmed to see if you want to post here.

There was an "above the fold" locked announcement for the BAUT forum's map for a while, is that what you are thinking of? Again, it's just my opinion, but these do take up extra space, and as with sections filled with a lot of stickies, they get very easy to ignore.

As far as the BA's and Fraser's support for the team, well how could we ask for more? They've given us privileged space here, on the best forum of its kind <suckingup>, several generous mentions in the UT newsletter, and now a sticky thread. To state the totally obvious, the team wouldn't exist without Fraser and Phil.

But the BA and UT websites, in addition to supplying top-class astro info, exist to support the whole BAUT enterprise. People need to buy those T-shirts and books they're hawking! For me, a blurb or link for a distributed computing team would probably distract from that. Just my opinion, like I say.

[Charlie in Dayton]

Oooookay...the SETI client is running, but apparently BOINC is suspended, and I don't know how to...unsuspend?...it...

I keep getting the message that the SETI client setup will cease operations on the 15th of the month. BOINC is installed (I have the little 8-ball icon down in the corner), but I haven't the foggiest notion of how to switch from SETI and make BOINC combobulate...

[Ken Vogt]

Hi CID,

The results for computer, here, shows your recent results, so you have been doing seti BOINC WUs successfully, indeed you've got over 1500 credits (cobblestones) under BOINC now.

Your last WU reported was on December 1, and since about that time seti has been having major problems with its upload and download servers. I have two seti WUs that have been trying for several days to transfer. That may be what has happened to the 4 WUs shown as "In Progress" -- they've all finished crunching, but BOINC can't upload them, or DL new work for you. In all probability, this is because of the very thing you mentioned: the shutdown of seti classic. (See my heartfelt appreciation of this project a few posts back. )

If you double click the "8" tray icon (perhaps it's a "B" ) you will open the BOINC manager, which shows what's going on. If you have any seti WUs running, they should show up in the "Work" tab there. The "Transfers" tab may show some WUs that are in the process of being transferred, as I say seti servers are overloaded ATM, but BOINC will automatically retry the transfers at random intervals.

If something else seems amiss other than no work being UL/DLed, you can look at the "Messages" tab and post here anything suspicious.

Anyhow, since you are running BOINC successfully, all you have to do is shut down seti classic & remove it from your start up group (or maybe just uninstall it.)

What most BOINC people do when projects like seti are down is just attach to another project and set it to run at a low priority, so that it only does real work when seti is down. predictor@home is good for this, as is LHC@home, because their WUs are fairly short.

If you want to try this, you can click Projects|Attach to Project from the BOINC manager main menu. A wizard will open (if you have BOINC version 5.x) and you just enter the project's url. The wizard has a link to a list of BOINC projects, and the BAUT team's projects are also listed on the BAUT team page linked in my sig (except CPDN, predictor, and µFluids.) If you do attach to another project, you can then join the BAUT team for it, which you can do near the bottom of the team page.

The whole BOINC thing is a bit intimidating at first, so please post back if you need anything else, excluding as always free beer.

But to summarize, it appears you are BOINCing just fine, and are just being caught out by a hopefully temporary bunch of overloaded seti servers.

[Laser Jock]

After months of begging, I finally have a new computer at work. I needed the computer to do some of the more complex calculations that my advisor wanted me to do. However, when it's not working on my physics, it can work on Einstein @ home! It's a Athlon 64 4000+ with 4 GB of RAM so it can chug through a WU in ~4.7 hours. So to those above me, watch out. I estimate my RAC will soon be close to 500.

[Ken Vogt]

Nice little box, Laser Jock!

Measured floating point speed	2487.8 million ops/sec
Measured integer speed	        4658.7 million ops/sec

I'm suitably envious.

You are already doing pretty good, becoming the most recent in E@H to go over 50,000 credits.

Wolverine will be next momentarily, making 15 BAUTers at that level.

---
Speaking of Einstein@Home,

Yes, S5 has started, with a lot of work. Things seem to be stabilizing, though.

It will be a few months before Einstein@Home starts working on S5 data, though. We're treating S4 as a serious search, of course, but we're also using it as a guinea pig for some better analysis tricks we'll need for S5. Also, there is a lot of work that goes into cleaning up the data and cutting it up into bite-sized chunks that won't kill everyone's computers and internet connections.

Stay tuned,
Ben

This S5 data set is the first for LIGO at its design sensitivity, IIRC, and the hope is for LIGO to soon detect gravity waves from inspiraling binary pulsars.

Hopefully the E@H search will also find a few of its target, single pulsars with mountains, as Ben Owen describes them.

So let's everyone on the team go out and pick up (or have Santa pick up for us) a couple of machines speced out like LJs...

[azazul]

So let's everyone on the team go out and pick up (or have Santa pick up for us) a couple of machines speced out like LJs...

Although its not quite up to those specs, I have ordered a replacement processor for the computer that went down several months ago. Mercury should be up and running within a week.

[Wolverine]

Whee, I finally broke the 50k mark.

[azazul]

Congratulations Wolverine.

[Laser Jock]

Whee, I finally broke the 50k mark.

And passed me too. Not for long, I've a pretty long list of pending WU's building up.

[Wolverine]

Thanks. And, yes, I'm sure it's temporary given my single, slowish box.

[azazul]

We now have a team in Rosetta@Home.

[Wolverine]

On E@H, we crossed the 2,000,000 mark for total credit!

(2,028,513.67 at the time of this posting)

[azazul]

On E@H, we crossed the 2,000,000 mark for total credit!

I knew we were coming close, but I forgot to check it recently. Also of note is that our recent average credit has been growing recently. At least partially thanks to Laser Jock's new computer. But some others seem to have been generating more credit recently, too.

[Wolverine]

I knew we were coming close, but I forgot to check it recently.

Same here.

[Laser Jock]

I knew we were coming close, but I forgot to check it recently. Also of note is that our recent average credit has been growing recently. At least partially thanks to Laser Jock's new computer. But some others seem to have been generating more credit recently, too.

I can't take full credit. My RAC has only risen by about 100. However, it is still rising.

Unfortunately, we were just passed by our old friend US Navy.

[mrkami]

But some others seem to have been generating more credit recently, too.

my RAC has gone up, I haven't a clue why though, a few week ago I was above 50 now I'm 19.

I've been having problems wth the site tonight, I can't seem to access my account page or the forums, is it just me?

[azazul]

my RAC has gone up, I haven't a clue why though, a few week ago I was above 50 now I'm 19

When you first join, you don't have very high RAC for a while. BOINC figures your RAC by finding the average credit you turn in each day with more weighting on recent days. So it takes about a month to build up much RAC.

I've been having problems wth the site tonight, I can't seem to access my account page or the forums, is it just me?

Works fine for me now.

[Ken Vogt]

News from LHC@home:

19.12.2005 12:21 UTC
CHRISTMAS AND FUTURE PLANS:

Just received from the LHC accelerator physicists:
"The results from the recent tracking campaign have been presented in our meeting last Friday. The presentation was well received and it was decided to continue with another massive study after Christmas, using a slightly different set of parameters. The foreseen study will need about the same computing resources as the one just completed. Maybe more, probably not less. The results of the previous study will be presented at the forthcoming Particle Accelerator Conference 2006 in Edinburgh."

Happy Christmas and thanks from CERN to all LHC@home volunteers!

And from this Einstein@Home thread:

We've probably got several months of crunching left on S4 data. Before too long you'll be seeing a new type of S4 work units which are basically analyzing longer stretches of data. (Don't worry, the individual work units will take about the same time to crunch as they do now.)

We really want to analyze longer stretches of data, but this requires some extra tricks. Some of them are associated with practical problems of not inflating the work units and killing everyone's machines when they run the screensaver. More fundamental is the fact that we have to start taking into account the intrinsic frequency drift of the source. Rather, all possible frequency drifts measurable in whatever time a data stretch lasts. This presents a tricky mathematical problem. We think we've solved most of it, but there is still some trouble with the validator and Linux machines. (My Linux machine, actually.)

Once that is sorted out (which will be soon), you'll be getting these new-style S4 work units for a few months. If all goes well, we should be shipping out S5 work units around the middle of next year. (S5 just started a few weeks ago, and it takes a lot of work to clean up the data, pick out the good stretches, and cut them into work unit-sized pieces.)

(bold added)

------
Many thanks to azazul for starting the Rosetta@home team.

December 9, 2005
Details about how we are applying Rosetta to help find cures for diseases such as HIV/AIDS, Cancer, Malaria, and Alzheimer's can be found in our new Disease Related Research section.

I know nothing about biology, and now prefer rosetta to predictor solely on the basis of the information in the link just above. If you have a different view, please post here, or on the rosetta thread on azazul's forum. Obviously, any BOINC project can be discussed there also; guests welcome.

On the BOINC side, rosetta WUs seem quick, finishing in about 4 hours CPU on a ~2.8 GHz box, and they have generous 3 week deadlines, so it is a good project for slower machines, and for "filling in" for projects which are occasionally down, like LHC and seti.

------
On the credit front, modesty does not prevent me from saying that our total RAC is several hundred lower than it would have been because I've been doing the CPDN spinup. As of now my model is 20% done and has earned about 10,000 credits for the alpha project's "BAUT" team. Carl has said he will eventually add this credit to the real CPDN totals, but CPDN culture militates against discussion of credit on their boards -- everyone fancies themselves far more concerned with finishing climate models, and "credit whore" is the dirtiest of words. But I promise on behalf of the team (and on my behalf, whore that I really am) to remind Carl of this before the end of my run, some 3 months hence.

------
Here's a snapshot of the current team performance, now with credit in all the major BOINC projects:


Finally, Congratulations to team founder azazul upon his recent graduation from San Angelo University with a degree in physics! I believe this proves wrong the contention by advocates of ID that extremely improbable events can never happen.

[Wolverine]

Welcome to the forum, mrkami. And, congrats, azazul!

[Ken Vogt]

Nature news item on the move of seti@home to BOINC. (free)

As I wrote on 11/17,

Hopefully the press will make note of the occasion, and take the opportunity to explain BOINC's role as seti's successor.

From the article:

The boost in computing power may or may not bring SETI closer to its goal of making contact with intelligent life out there. But [David] Anderson notes that BOINC brings much needed help to intelligent life on Earth, for solving problems closer to home.

If you already have SETI@home or want to start contributing, you will have to download and install software from BOINC. You will then be given the option to donate various percentages of your machine's time to any or all of the projects. Users can chose, for example, to donate 70% of their unused computer power to finding curing disease, 20% to looking for aliens, and 10% to predicting the future climate.

I have "Curing Disease," caused by sitting too close to the woodstove on a winter's morning, so locating it will be a big help.

Seriously, it's nice to see reasonably accurate descriptions of BOINC in this magazine; though giving only 10% to CPDN is not realistic, given sulphur models' length.

[Udo]

So let's everyone on the team go out and pick up (or have Santa pick up for us) a couple of machines speced out like LJs...

Hello,

at my office I got a nice replacement server...
It's only a 2.8 GHz Intel but it's a Xenon dual processor system! Currently it delivers nearly exactly 6 Einstein WUs per day (that is 4h per WU).

And due to the end-of-work at LHC an other Server is currently crunching Einstein WUs...

overall my RAC climbed from approx. 850 to 1250 (with approx. 3000 Credits pending).

BUT...
when LHC delivers work again, one of my computers will return to those WUs...
and I have to shut down an other server (the Pentium III 1.4 GHz, which is replaced by the new one), which returned 2 WUs per day...

--------------

I also joined the Rosetta team as I was already crunching there for a while...
a long deadline, no quorum required... just the WUs for my home computer which is on for only 20%...
unfortunately they have big data files to download (2 MB) for only a short computation (1-4h depending on the WU). For my low speed modem connection at home that is not a good relation...
So I started to crunch a Einstein WU at home after pausing Einstein for a while... and the WU was still taken from the old data file (downloaded on 26th of september!)

so far from me, happy crunching and happy christmas!

Udo

[Ken Vogt]

Udo,

Sounds like a nice new upgraded machine; I'm suitably envious.

And welcome to the Rosetta team, it's good to have company there. I apologize for not noting the size of the WUs, with DSL you forget to pay attention any more, but 2 MB WUs would be a disadvantage for sure on dial-up or machines short on disk space.

Thanks for pointing it out.


___

December 23, 2005

After almost five months of development and private testing, Einstein@Home will begin public testing and distribution of a new search application called "Albert". This is an improved and more sensitive version of the previous search code. Initially we will send out only a few hundred workunits for this new search, but if these are successful, we will begin broader distribution of these new workunits over the coming few days. If unexpected problems arise, please be patient, and look in the Problems and Bug Reports message board for additional information.

These new WUs can be identified because they start with "z1_" (edit: or "r1_; thanks to Udo below) instead of h or l; and they use a new application; for Windows, your version will change from 4.79 to 4.37.

I haven't got one of the new WUs yet, and there don't seem to have been any problems yet reported on the boards.

As I understand it, the improved sensitivity is compensated for by a narrower range of frequencies searched per WU, so that crunch time will remain about the same. (edit: or less, see below.)

___
I join Udo in wishing the whole BAUT team and community a Happy Christmas and Holiday Season --- Greetings!

[Udo]

These new WUs can be identified because they start with "z1_" instead of h or l; and they use a new application; for Windows, your version will change from 4.79 to 4.37.

Ken,
I already got some of the new WUs.
In my case they start with "r1_" and take 27800 seconds to complete instead of 40800 (where my old WUs started with "w1_").
See WU 3080064 for more information...

Udo

[Ken Vogt]

Cool, Udo!!

I'll edit above to include the "r1_"

Dr Allen has a post about Albert, which he is updating with answers to crunchers' questions, so you only have to read the first post.

- The new WU have different execution times, typically ranging from about 25% to 100% the previous execution times

- The new WU application incorporates all BOINC graphics and other bug fixes to date

- The new WU application has a slightly re-arranged screensaver, which includes our top wish-list item: a real time clock

So apparently, contrary to what I said above, -oops , WU times will vary, but will often be shorter, as Udo found.

Who gets what WU is random for the moment...

[Fortunate]

I downloaded the latest version of BOINC that was available to me when I openend my accounts in September 2005. That was BOINC version 4 point something. I have never upgraded to a later version of BOINC because I heard rumors of "problems." I have a pretty fast computer, an
Athlon XP 3000+ with Windows XP. I only run LHC and Einstein. Should I upgrade to the latest version of BOINC? Are there any particular advantages or disadvantages to doing so?
Thanx.