LHC@home

I haven't posted since LHC@home started up again. I've been merrily crunching along since 2005! Within the last year I upgraded my old machine to an i7 950. For the last month or so every thing was going smoooooth. This morning I had 18 WU to crunch. As one would finish another would down load, just like clockwork! AND THEN--- no tasks sent---no tasks sent---no tasks sent. That has happened before during all the upgrades being made at your end but never lasted all day long. I also know that an i7 950 is capable of sse2,sse3, sse4,8,13,32,95/ just kidding but I never get any, how come. Hope all these things get worked out. Pick

---

Well, while puttsing about I saw my system start to download. As fast as they did they were done. Then the download errors started. I clicked on the transfer button and a half dozen were hung up in there. One click and they downloaded and just as fast they were gone. 30 or so in about 15 sec. .25% download errors. What's up? Pick

---

I gst no tasks since yesterday evening.
And 3 tasks hang in download than they become download error.
Greetings.

After seeming to resolve itself last weekend, this problem started up again yeserday, and has been unable to get work since through today. This isn't just 7.0.28, as beings that I've got a new laptop scheduled to arrive in 2 days (only ever built desktops, not laptops) I haven't bothered to update software on a system that's getting replaced/tossed out. This is occuring on 6.12.15 on Windows 7 Pro x64 also....

14,000 + available in queue, no work sent. There's also not much in the queue besides as I've been cleaning the queue out, to wind stuff down. They say tomarrow, but Fedex hasn't shown it yet, so I'm planning on 1 or possibly 2 days left to crunch (I'll check the trackinng latter), but it doesn't take that many hours to crunch these, so....

---

After getting a few WUs, I am now back to zero WUs but there
are ~14k WUs available to send. Whatever it was - it's back...

---

I think I am back to normal now. Happily crunching 4 tasks at the moment. It's just been a heck of a week for everybody!

I never had any doubt that the problems would get resolved, and I don't worry about loss of credit because it's all for science anyway. And such great science it is. I'm with this project all the way.

Strange that after stopping LHC for a week or two due to the wrong applications being downloaded and Linux being so slow compared to Windows, that when I allow tasks I am unable to get any at all.

Server says over 8,000 WUs ready to send.

Keeps saying "Got 0 new tasks", "No tasks sent".

Tried a reset and a detach and reattach but still the same, Linux and Windows.

Oh well I did finally get to the 50,000 milestone the other week, guess I wont be getting to the 100,000 milestone any time soon, never mind I gave it a go.

Conan

---

Yup, me too. What was working no longer works... sigh... not even the i7-950 box gets any more tasks...

Have to smile about that! It's a sure sign folks are tweaking the system.

LOL!

:)

Woopsy!

Just got a load of (pni) work on the compaq...

Running now!

:)

Allowed work fetch again and this time work downloaded.
PNI work units on 3 machines and SSE2 on the other. A number of work units downloaded on the Windows machines but only 1 each for the Linux machines.
I am interested to see how they run as the Linux WUs before were way behind the 8 ball.

Conan

---

Just got another load on the i7-950 system... so looks like things moving along well.

This load was SSE3 and PNI both...

On the Compaq box (running Linux at the moment), the tasks take 4 hours instead of the usual 7 or so... so much improved.

I LIKE what they are doing! The tweaks are improving things all around!

Hip Hip Hurray for smart developers! I bet this project get tuned perfectly very soon!

:)

interesting.

There should be no difference between linux and windows for the #tasks downloaded. Of course, there is an influence of the past results. So, if the linux boxes were much slower, their average turnaround parameter is higher and they are considered "less reliable" by the system and therefore get less work.

This should correct itself over time, as new results come in.

Allowed work fetch again and this time work downloaded.
PNI work units on 3 machines and SSE2 on the other. A number of work units downloaded on the Windows machines but only 1 each for the Linux machines.
I am interested to see how they run as the Linux WUs before were way behind the 8 ball.

Conan

---

skype id: igor-zacharov

Okay, have the new Linux installed on the 800D box with the Sabertooth Mobo... had to update the Realtek driver to get the LAN to work right and now it's humming fine.

It has also downloaded only PNI tasks and they are running fine.

I like running linux since many projects run significantly faster using Linux. For instance, Correlizer@home tasks take 11:35 under windows on this box, but the same tasks take only 9:54 under Linux. Same for many others..

Thanks for all your hard work! Looks like all my systems talking and working well now!

And the SSE3/PNI tasks run almost twice as fast... 7+ hrs old tasks vs. 4hrs SSE3/PNI on the 1055T box (3.25GHz) for instance.. WOOHOO!

:)

We've got a lot of work to do! Nice to see 50000 tasks in progress and an adicional 20000 to send.

Congratulations.

---

We've got a lot of work to do! Nice to see 50000 tasks in progress and an adicional 20000 to send.

Yes, indeed... Great to see the steady work. Keep it coming! :-)

---

It didn't last long ... they have been all distributed.

Well I am doing my best,but we have some limitations on the
user side at CERN. Also we have to introduce some new elements
(physics) in he next few days, generate an sse4 executable, fix
the MAC executable etc etc.

However, since "you" are delivering the equivalent CPU capacity
of the entire CERN computing centre, but really "only" 50%, since
we run every task twice, I shall make a further effort.
As suggested here on the Message Boards I am submitting work in batches.
I shall try increasing the number of batches.

Incidentally, Igor has just reported some interesting numbers which I
hope he will pass on. Basically, if I remember well, we are rejecting results from
something like 50 hosts out of 25,000 (he will correct), which means I am rather
confident about numeric portability. I am also planning to send a test to these only 50
volunteers. In fact, Igor had the idea of including a test of a few turns in each
Work Unit. If this worked I could then run each case only once, with
perhaps a few random duplicates, in order to double the capacity! :-)
Well it is perhaps a dream. This would be easier than programming GPUs
as SixTrack is Fortran.

However, the main thing at the moment is to document, maintain the quality
of service to both users and volunteers, and get on with the studies of the
various LHC upgrade options. Eric.

---

Well I am doing my best,but we have some limitations on the
user side at CERN. Also we have to introduce some new elements
(physics) in he next few days, generate an sse4 executable, fix
the MAC executable etc etc.

However, since "you" are delivering the equivalent CPU capacity
of the entire CERN computing centre, but really "only" 50%, since
we run every task twice, I shall make a further effort.
As suggested here on the Message Boards I am submitting work in batches.
I shall try increasing the number of batches.

Incidentally, Igor has just reported some interesting numbers which I
hope he will pass on. Basically, if I remember well, we are rejecting results from
something like 50 hosts out of 25,000 (he will correct), which means I am rather
confident about numeric portability. I am also planning to send a test to these only 50
volunteers. In fact, Igor had the idea of including a test of a few turns in each
Work Unit. If this worked I could then run each case only once, with
perhaps a few random duplicates, in order to double the capacity! :-)
Well it is perhaps a dream. This would be easier than programming GPUs
as SixTrack is Fortran.

However, the main thing at the moment is to document, maintain the quality
of service to both users and volunteers, and get on with the studies of the
various LHC upgrade options. Eric.

---

Great news Eric, Thank You!

Now imagine, if you send an EMail to all users (or all inactive last 30 days) and only half of the currently inactive are coming back.......

I guess you should increase the available discspace before that message to not run out of it again ;-)

AND you get it to produce a GPU app. (Dreaming on? Or not?)

---

Christoph

Well I am doing my best...I shall make a further effort...I shall try increasing the number of batches.

Thanks for all the hard work, Eric. I'm just glad to see the significant increase in work recently and happy to crunch away whenever the tasks are available. :-)

Cheers,

MarkR

---