What is happening at CERN Team?

[TPL]

I made observation when checking my own last CPUs page. There have been constantly the same user with unsuccessful WUs.

Check for example these two:
https://apps.foldingathome.org/cpu?q=ALICE-IPNL
https://apps.foldingathome.org/cpu?q=ALICE-Legnaro

What is the point to go on like this? Huge load to servers.

I checked some of those WUs and former one has some successful WUs but latter none that I could find. Still there are over 23000 lines on that page.

[NRT_AntiKytherA]

Maybe the machines associated with those IDs have stability issues they are not aware of yet. Given the fact most of the world has been on lockdown in the last 3 months it is entirely possible that nobody has been able to do anything about this even if they are aware of any such issues. Lockdown is beginning to be lifted at differing rates so I wouldn't be too concerned about this.

[TPL]

It seems there are many of those user accounts failing the most if not all of WUs. It can make 100's of thousands of wasted downloads all together and it seems to happen frequently, every 1-10 seconds.

Also there are a huge number of clients. Like every time new ID. I don't know why but was wondering if they know of this big failure rate. Maybe they should.

I think CERN computers are possible to manage remotely anyway, lockdown or not, but what do I know about it.

[WTS]

It seems there are many of those user accounts failing the most if not all of WUs. It can make 100's of thousands of wasted downloads all together and it seems to happen frequently, every 1-10 seconds.

Also there are a huge number of clients. Like every time new ID. I don't know why but was wondering if they know of this big failure rate. Maybe they should.

I think CERN computers are possible to manage remotely anyway, lockdown or not, but what do I know about it.

When the Pandemic is over, and we have both a vaccine and a treatment, it'll be interesting to see how much of our efforts meant something, and how much of our efforts were wasted. On the subject of folding in general, I hope we'll have taken some steps into the future, and I hope that many will stay with us. We'll always need more folders, even after the Pandemic.

[NRT_AntiKytherA]

All rather depends on whomever usually manages them being in a fit state to do so or have the means to at home. If you are that concerned try emailing CERN about it

[TPL]

When the Pandemic is over, and we have both a vaccine and a treatment, it'll be interesting to see how much of our efforts meant something, and how much of our efforts were wasted. On the subject of folding in general, I hope we'll have taken some steps into the future, and I hope that many will stay with us. We'll always need more folders, even after the Pandemic.

I think we will never receive any thanks for a good job and we still don't know if any vaccine will be effective. But better try than not. There is no known vaccine for any corona virus atm. That is one of the most important reasons to keep going. There will be more than this one. Actually thousands. So lets fold on. I find no reason to stop when this will be over. You know, Ebola for example is still alive and kicking. Cancer is not gone anywhere nor Alzheimer. There is a lot to do and will be that way.

Anyway, this kind of WUs continuous failing right in the beginning will not be helpful at all but only waisting resources. A lot of bandwith there for nothing.

I AM a little bit worried when I find CERN is failing that much on just basic level. It should be one of the leading research bodies in nuclear research in the whole world with the pack of the best scientists.

But anyway, if this is ok to FAH-Team, it is ok to me of course. It would be a little embarrassing to notify them by me and I wouldn't like to be the one.

[NRT_AntiKytherA]

I don't know about anyone else here but I certainly don't fold in the hope of ever being thanked for it, the 'thanks' we will receive is not dying from the researched diseases ourselves or saving future generations from the same fate. I concur about the wasted bandwidth and hardware runtime though, it's not ideal for one normal machine let alone a rack of blades or two. They have essentially got security, building and services maintenance crews at CERN only right now.

https://home.cern/news/news/cern/gradua ... activities

[PantherX]

I don't know about anyone else here but I certainly don't fold in the hope of ever being thanked for it, the 'thanks' we will receive is not dying from the researched diseases ourselves or saving future generations from the same fate...

AFAIK, the research paper does acknowledge Donors (https://www-cs.stanford.edu/people/ihaq ... gpuser.pdf):
"We foremost thank the Folding@home donors, without whom this study would not be possible."

[Neil-B]

This has been reported to a CERN contact ... It may look "bad" (and those are just a couple of "users" with issues) and definitely needs resolving but do take into account the vast number of WUs CERN are folding - check EOC top ten folders - they had 3/4 in the top ten recently so what looks bad is actually a small percentage against the good they are doing.

[NRT_AntiKytherA]

I have seen this in research papers too and that's nice recognition but I still see the potential cures as the real reward

[TPL]

I have seen this in research papers too and that's nice recognition but I still see the potential cures as the real reward

Me too. As I said I'm not expecting any more. It would be nice to know how actually we have helped but not excpecting to ever know.

Please understand my concern about server heavy loading on this. What we see from last CPUs page is the last WU for every ID. We don't see any previous. Seems like many of those ALICE-coded "users" have had problems and we don't know how many times per ID in the past. In worst case this can make millions. If there are 23000 lines all failures how much it can be in total during time?

I'm happy it has been informed.

If you see something like ALICE-ZA_CHPC, Credit 8,945,842, WUs 267,522 it makes 33 points/WU avg. Can't be true and it isn't. There are not just couple of those "users".

Also these failed units are increasing points and WU count but not doing any good. So there is a reason to lift this up. I loaded one all-failed-page into LibreOffice and sum of it's points is 285k. 23k WUs.

[mlassnig]

I've notified my colleagues at CERN who are running the F@H clients on our distributed infrastructure. It's past midnight now, so I can't guarantee that somebody will be able to do something in the next few hours though.

[mlassnig]

Hi, my colleagues told me that it is difficult to say from their side what the actual issue is with these nodes (as they run our jobs just fine), therefore they have turned the F@H clients off on them so it should not impact you any more.

[toTOW]

If they want to post a couple of logs here, we could help them find out what's wrong ...

[fwyzard]

Hi,
if anyone ever sees issues with the contributions from CMS-Experiment, please do contact me - either PM me, or email me.
I assume the admins can access my email form the forum registration and/or from the passkey registration.

.Andrea