1) Message boards : Number crunching : GPU advertised for LHC, but they don't do it? (Message 41995)
Posted 4 days ago by Peter Hucker
Post:
https://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html
This might be a good starting point to understand why it is not a good idea to work with reduced precision.


I know reduced precision is bad. I don't have enough maths skills to understand any papers on it. But can't more calculations be done at FP32 to make it more accurate? Even if it requires several, it would still be faster than FP64 on most cards.
2) Message boards : Number crunching : GPU advertised for LHC, but they don't do it? (Message 41991)
Posted 4 days ago by Peter Hucker
Post:
Well, 64 bit floats still give you rounding errors but they tend to be way out there past the decimal points. :)


Can more calculations at FP32 not be done to make the accuracy better? On cards with stupidly low ratios like 1:24, Even doing 10 calculations at FP32 would be faster than 1 at FP64.
3) Message boards : News : CMS@Home up again (Message 41695)
Posted 22 Feb 2020 by Peter Hucker
Post:
Today I have been getting only Theory on my main cruncher. It has been now accepting work from all subprojects except sixtrack. Yesterday it had all subprojects selected and it got only sixtrack tasks. So no CMS for me although it has also been selected for a couple of days now.


Seems like the server is picking favourites :-)

Sixtrack is very short of tasks, server status usually shows 0 available. One of my computers managed to grab 15 of them last night, but that's all. Sixtrack is the only one that will work without virtual machine, so anyone can do it, including mobile phones (which I have two of), and my three antique computers with very old processors and small RAM.
4) Message boards : News : CMS@Home up again (Message 41693)
Posted 22 Feb 2020 by Peter Hucker
Post:
The search function might be your friend since this has been explained a couple of times in this message board.

The server fills it's ready to send queue from a couple of upstream processes each representing one of the subprojects.
Now the server's shared memory holds a list of "results" in random order.
In addition large projects like LHC@home spread the load over a couple of servers which are contacted in random order (DNS based load balancing).

Your client generates a request to get x seconds of work and the server that answers your request will send you the n first "results" from it's shared memory list.
- n is calculated based on the sum of the estimated runtimes.
- server side quotas will be respected.
- results from deselected subprojects will be skipped*).

Under certain circumstances this leads to a situation where one of the servers has no tasks (=result) from your active subprojects in it's queue and you will get a "no tasks available" message although the server status page show lots of available tasks.


*) This might lead to the situation that the next client who has this subproject checked will get the skipped "results".


I suspect that last point is why I only get CMS. A lot of folk have probably turned CMS off due to the problems, and I can see there are less "users in last 24 hours" on the server status page. Hence CMS is probably the first 50 tasks in the queues. I ain't turning my CMS off. I don't care if I get no credit, if it helps them sort out the problems, my computer will try to do them. Some of them work.
5) Message boards : News : CMS@Home up again (Message 41691)
Posted 22 Feb 2020 by Peter Hucker
Post:
I'd like to know how LHC servers (and other projects) decide what subproject to give you. If I have them all selected, I could understand getting the one with the biggest queue, or maybe first in first out, but with LHC at the moment, I got given loads of CMS and no Atlas or Theory, despite CMS having the least jobs available of the three. Maybe they prioritize one over the other? Maybe if very few people have CMS enabled, those that do just get CMS?
6) Message boards : News : CMS@Home up again (Message 41686)
Posted 22 Feb 2020 by Peter Hucker
Post:
Most of my CMS are causing errors again. I'm assuming this is a CERN fault and not my doing. Please let me know if I can adjust anything at this end. Running latest Boinc and Virtualbox under Windows 10.

And I'm still not getting Atlas or Theory tasks, despite there being more of those showing as available on the server status page. I'm only being given CMS.
7) Message boards : Number crunching : No tasks at all available? (Message 41684)
Posted 22 Feb 2020 by Peter Hucker
Post:
Most of my CMS are causing errors again. I'm assuming this is a CERN fault and not my doing. Please let me know if I can adjust anything at this end. Running latest Boinc and Virtualbox.
8) Message boards : Number crunching : No tasks at all available? (Message 41681)
Posted 21 Feb 2020 by Peter Hucker
Post:
We had an outage of some IT services at CERN yesterday, so task submission for some applications stopped for a while. Back to normal today. Hopefully SixTrack users will submit a new batch soon too.

Thanks for your contributions and happy crunching!


I have three older computers and two phones that can only manage sixtrack, I hope they come back soon, I only get the odd one (perhaps an aborted task recycled?)

My decent computer is set to take any task, but it's only being given CMS, despite there being more of Atlas and Theory shown on your server status. Am I getting only CMS because a lot of others still have CMS switched off?
9) Message boards : Number crunching : No tasks at all available? (Message 41675)
Posted 20 Feb 2020 by Peter Hucker
Post:
I've got 4 CMS tasks that have run for quite some time without failing, so I think they might have fixed everything :-)

That would be useful, if they ever get it really fixed. You never know.


Well the others were failing in between 10 and 40 minutes. These 4 have now been running for over an hour.
10) Message boards : Number crunching : No tasks at all available? (Message 41673)
Posted 20 Feb 2020 by Peter Hucker
Post:
I just downloaded a native ATLAS.
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=133600650


Yip, server status now shows a small number of everything available except Sixtrack (which is unfortunate as all but one of machines can only run that, they can't do virtualbox as they're not powerful enough.

I've got 4 CMS tasks that have run for quite some time without failing, so I think they might have fixed everything :-)
11) Message boards : News : CMS@Home -- ongoing problems (Message 41670)
Posted 20 Feb 2020 by Peter Hucker
Post:
Note: my computer is currently processing 10 CMS tasks which are erroring due to the server problems. Is that causing me to get limited in downloading more tasks?
The maximum number of tasks is twice the number of the cores.


It's a 6 core machine, although only using 4, as the other 2 are to assist GPUs on Einstein. Not sure which number it takes, but presumably the 6 as I had 10 CMS at once.

It's completed all the CMS tasks now, so I have 0 tasks, but it won't take the Theory ones that are available. I assume there's either:

A problem with the LHC Network outages:
https://lhcathome.cern.ch/lhcathome/forum_thread.php?id=5072&postid=41658
https://cern.service-now.com/service-portal/view-outage.do?n=OTG0054949
https://cern.service-now.com/service-portal/view-outage.do?n=OTG0054961

Or the server is upset about my CMS tasks erroring, so presumably I'll get tasks shortly.
12) Message boards : Number crunching : No tasks at all available? (Message 41669)
Posted 20 Feb 2020 by Peter Hucker
Post:
Thanks, my computer will come back whenever they're ready. It can busy itself on other projects just now.
13) Message boards : News : CMS@Home -- ongoing problems (Message 41665)
Posted 20 Feb 2020 by Peter Hucker
Post:
I can see 1254 tasks available for Theory in the server status, although unusually Atlas has run out.

But my computer is requesting tasks and not getting those Theory ones.

Is something up?

Note: my computer is currently processing 10 CMS tasks which are erroring due to the server problems. Is that causing me to get limited in downloading more tasks?
14) Message boards : Number crunching : No tasks at all available? (Message 41664)
Posted 20 Feb 2020 by Peter Hucker
Post:
I can see 1254 tasks available for Theory in the server status, although unusually Atlas has run out.

But my computer is requesting tasks and not getting those Theory ones.

Is something up?

Note: my computer is currently processing 10 CMS tasks which are erroring due to the server problems. Is that causing me to get limited in downloading more tasks?
15) Message boards : News : CMS@Home -- ongoing problems (Message 41662)
Posted 20 Feb 2020 by Peter Hucker
Post:
I'll be leaving my main computer doing LHC. If an error helps, then it's worth doing.

Unfortunately only 1 of my 4 computers can run LHC. The older 8GB RAM machines (my main one is 16GB) with rubbish processors seem to screw up when doing any Virtualbox stuff. They always give way too many computation errors on the tasks. They do have virtualisation capabilities which is switched on in the BIOS, and Virtualbox is installed correctly in the same way as the main machine. Oh well, they're on Rosetta and Universe instead.
16) Message boards : News : CMS@Home accidentally shut down -- Please set No New Tasks (Message 41567)
Posted 14 Feb 2020 by Peter Hucker
Post:
You could email everyone if there's something urgent - we all have the option to tick (or not) the thing that says it's ok for admins to email us. Then we'd see it quicker.

Are you aware that you can subscribe to MB threads or get them via RSS feed?
The way this works can be configured on the MB pages and on this page:
https://lhcathome.cern.ch/lhcathome/edit_forum_preferences_form.php
It's far more than simply allow an admin to send you a mail.


Sorry, I have no idea what the point of RSS is and have never felt the need to use it. Everything I use emails me when something needs my attention (eg when you replied in this conversation). I really don't want yet another source to check and another program to run. In my life I have four inputs - phonecalls, texts, emails, and physical post (mail). I don't want five.
17) Message boards : News : CMS@Home accidentally shut down -- Please set No New Tasks (Message 41552)
Posted 12 Feb 2020 by Peter Hucker
Post:
"My smaller three desktops produce a lot of errors when running virtual machine LHC (they're only 8GB RAM with old processors)" 4gb ram on 9 year old cpu still working fine for me with LHC.


Which subprojects does it run? I didn't bother selecting, I let them do anything. I got computation errors on about 75% of them, so thought it a waste of time.

And what CPUs do you have? One of mine for example is a Q8400 with 8GB DDR2.

And what OS? Mine all run Windows 10.
18) Message boards : News : CMS@Home accidentally shut down -- Please set No New Tasks (Message 41548)
Posted 12 Feb 2020 by Peter Hucker
Post:
Although couldn't the Boinc server have been told not to hand out any CMS?

It sometimes happens that the BOINC server doesn't immediately stop sending task envelopes, especially when the subtask shortage is not very long.
In this case Ivan sends out an NNT info.
This ensures a faster CMS restart when fresh work is available.


Ok, although chances are most people won't see that message for a day or so. I usually don't even notice them at all.

Yeah, sure, but I have to do my best. The problem is that when no jobs are available, then the BOINC task will return an error. Your BOINC manager will flag that as an error, and reduce the quota of tasks you can request per day, until this gets down to unity. The quota will only increase when you run a successful task. So if you don't set NNT then your machine gets reduced to one task request per day, until new jobs are available, allowing a task to complete, thus increasing the per-day quota. So, it takes several days to ramp up to full production again if you allow your machine to deplete its quota.


Ok, thanks for explaining in detail. I take it some people run only CMS on their clients? I have Androids that only do Sixtrack (that's all that gets given to them, I assume Androids can't do the others), but my main desktop does any LHC task, and rarely gets CMS, presumably since Theory and Atlas have a lot more tasks available. My smaller three desktops produce a lot of errors when running virtual machine LHC (they're only 8GB RAM with old processors), so I don't use LHC on those.

You could email everyone if there's something urgent - we all have the option to tick (or not) the thing that says it's ok for admins to email us. Then we'd see it quicker.
19) Message boards : News : CMS@Home accidentally shut down -- Please set No New Tasks (Message 41535)
Posted 11 Feb 2020 by Peter Hucker
Post:
Although couldn't the Boinc server have been told not to hand out any CMS?

It sometimes happens that the BOINC server doesn't immediately stop sending task envelopes, especially when the subtask shortage is not very long.
In this case Ivan sends out an NNT info.
This ensures a faster CMS restart when fresh work is available.


Ok, although chances are most people won't see that message for a day or so. I usually don't even notice them at all.
20) Message boards : News : CMS@Home accidentally shut down -- Please set No New Tasks (Message 41529)
Posted 11 Feb 2020 by Peter Hucker
Post:
Your client requests a task from the BOINC server. That task creates a CMS-VM on your machine and that VM requests CMS-jobs from another server and that's out of jobs or unreachable.


Ah that makes sense. Although couldn't the Boinc server have been told not to hand out any CMS?


Next 20


©2020 CERN