21) Message boards : ATLAS application : queue is empty (Message 48616)
Posted 22 Sep 2023 by Mr P Hucker
Post:
If the percentage complete in Boinc is correct I'm doing 3 tasks at once on my Ryzen 9 3900XT, in 3 hours. They don't look excessively large to me. Atlas has always taken about that long. They're at 50%, so Boinc could of course be insanely wrong.

Edit: one finished in 2h11m (9h37 CPU).

It's been accepted as valid, although I notice the one I did 6 days ago on the same machine was 10 times longer, maybe this is a one off: https://lhcathome.cern.ch/lhcathome/results.php?userid=55945&offset=0&show_names=0&state=4&appid=14
22) Message boards : ATLAS application : queue is empty (Message 48614)
Posted 22 Sep 2023 by Mr P Hucker
Post:
I'm just going to run them. I've grabbed a bunch on my 5 fastest computers. Might aswell help out with whatever they're trying to do/test/fix.
23) Message boards : ATLAS application : queue is empty (Message 48610)
Posted 22 Sep 2023 by Mr P Hucker
Post:
Is the way to successfully complete these long units is to use more processors to accelerate completion time of the work unit? When does the "kill switch" appear, after about a day and a half of processing?

I just answered my own question. Both units were cancelled by the server at precisely 112,671.63 seconds.
I don't understand using anything less than the 8 threads the program will use. Why do twice as many at once in double the time?
24) Message boards : ATLAS application : queue is empty (Message 48608)
Posted 21 Sep 2023 by Mr P Hucker
Post:
They could have announced it on the forum and leave the decision to abort to the crunchers. They could also have tested them on the dev-site first. But it is what it is, there's no point in crying over spilled milk.
I wasn't aware Boinc had the facility to remotely abort a started task. Perhaps it's just other projects don't like doing so. I'm not sure why anyone would want to finish a task which is just going to be discarded at the project end.
25) Message boards : ATLAS application : queue is empty (Message 48606)
Posted 21 Sep 2023 by Mr P Hucker
Post:
If something changes so the task is pointless to continue, would you rather waste even more time or abort? It's not their fault they realised what they asked you to do is now incorrect.
26) Message boards : ATLAS application : queue is empty (Message 48603)
Posted 21 Sep 2023 by Mr P Hucker
Post:
Last one I did was 12th September (not sure if that's 2000 size?) took 21 hours on my Ryzen 9 3900XT. I estimate your slowest PC would take double that, just under 2 days. Sounds doable, what's the problem?

https://lhcathome.cern.ch/lhcathome/result.php?resultid=399275439

I shall put everything on Atlas here and see what happens, I have very slow computers and reasonably fast ones.

Great, changed all the settings and cancelled work from other projects then find there's none left to get. Well that was a waste of time.
27) Message boards : ATLAS application : queue is empty (Message 48564)
Posted 16 Sep 2023 by Mr P Hucker
Post:
I got a 2k job and let it run into an error cause the first event needed 787 seconds and the second 2800 seconds.
Let's say on average 1500 seconds times 2000 makes 3 million seconds = 1 million on my 3 threaded VM = 11.5 days 24/7.
Received ---------- 16 Sep 2023, 9:33:40 UTC
Report deadline -- 24 Sep 2023, 8:44:17 UTC
Make up your mind.
You would have made it on 8 threads. Why throttle the poor thing?
28) Message boards : ATLAS application : queue is empty (Message 48562)
Posted 15 Sep 2023 by Mr P Hucker
Post:
hi guys, I have a question?
From where do you download ATLAS tasks?
As per the Server Status page, there are none available ...
They appear now and again, they're just testing now. Should be a steady stream shortly. Set your account to get atlas only with other types if none available.
29) Message boards : ATLAS application : queue is empty (Message 48558)
Posted 15 Sep 2023 by Mr P Hucker
Post:
I selected 4 threads to keep the utilization high (= trying keep the time that only a single thread is used short). With these long tasks I probably will raise the thread count for single task to shorten the run time. Let's see how long these tasks actually will take.
If you're not getting full utilisation, just run more of them. For example, I have a 24 thread machine and tell four Atlases to run. I let them use up to 8 threads, but tell Boinc they use an average of 6. If you use Boinctasks, you can see what percentage they used and change it accordingly. Not sure if Boinc Manager does this as I never use that horrid interface. For example when I had them set to use 8 threads, it said 75% usage. 75% of 8 is 6, so I set them to 6.

My slower host only has 8 threads of which I leave 2 free and 2 are busy with Einstein GPU tasks, so that leaves only 4 threads for LHC.
GPU tasks don't always need a full core. I adjust and make sure the GPU usage stays high. You can also run more than one task per GPU so the GPU has something to do while waiting for the CPU.
30) Message boards : ATLAS application : queue is empty (Message 48556)
Posted 15 Sep 2023 by Mr P Hucker
Post:
Four Atlas tasks running here on two hosts, 4 cores used for each task. Run times are currently 48...66 hours (CPU times 156... 248 hours) and all 4 cores still active as seen on Top (console Alt+F3).

If you view the console Alt+F2 you can just make out the number of jobs finished so far (mine are at 1200...1800 finished) averaging about 2000 seconds per job. So they are taking forever to finish. :-(
I estimate your fastest machine is 3/4 of the speed per thread as my Ryzen 9 3900XT. So if you allowed it 8 threads, it would take 1 day compared to my 17.5 hours. With only 4 threads, I'd expect you to take 2 days, so they're fine. Any particular reason you only use 4 threads?
31) Message boards : ATLAS application : queue is empty (Message 48554)
Posted 15 Sep 2023 by Mr P Hucker
Post:
When you have in prefs all LHC-Projects active,
Boincmanager take the Project from Cern-IT they give us.

So, you can take default, home, school or work to make a difference of work for each of your PC's.
Yes, that's what I do, I have a venue for each type of task, and put a few machines in each.

All Projects together in one PC don't work really good.
I have no problem running them all on one PC, if I happen to get a selection.
32) Message boards : ATLAS application : queue is empty (Message 48552)
Posted 15 Sep 2023 by Mr P Hucker
Post:
When I set the max threads on LHC, all I got were Atlas tasks. I also want to run Theory tasks,
Yes, you never get what you ask for here. If I select to get any tasks, I only get CMS. Every other project gives you an even mixture. Since I have more than one computer, I just set some to get CMS, some Theory, some Atlas, but leave the "if nothing available give me something else" ticked. Then if everything is available, I get some of each. I have to do this as CMS maxes out my uplink and Theory maxes out my downlink. Half of each and it seems to manage without throttling the CPUs. If I had only one computer I've no idea how I'd get a mixture.
33) Message boards : ATLAS application : queue is empty (Message 48549)
Posted 15 Sep 2023 by Mr P Hucker
Post:
hadron wrote:
Are you overclocking your CPU? Or maybe running each task on multiple threads?
I have a Ryzen 9 5900X which has almost the same base frequency as the 3900XT, so I would expect your times and mine should be roughly equal.
Actually your CPU is 18% faster according to benchmarks.

Of course I use multiple threads, ATLAS is designed for this. I've left it on the default of 8 threads per task, it tends to average 7, so that's what I told the scheduler via app_config.

    <app_version>
        <app_name>ATLAS</app_name>
        <plan_class>vbox64_mt_mcore_atlas</plan_class>
        <cmdline></cmdline>
        <avg_ncpus>7.000000</avg_ncpus>
        <ngpus>0.000000</ngpus>
    </app_version>

I haven't overclocked.
34) Message boards : ATLAS application : queue is empty (Message 48547)
Posted 15 Sep 2023 by Mr P Hucker
Post:
hardon wrote:
Lucky you. I have 4 Atlas tasks running right now. They've been running for 2 days 10 hours, and still have 2 hours to go.
17.5 hours total to run here on a Ryzen 9 3900XT, unfortunately boinc has decided to run something else and interrupted it. I set them to stay in memory while suspended which really helps.
35) Message boards : CMS Application : no new WUs available (Message 48485)
Posted 2 Sep 2023 by Mr P Hucker
Post:
Just 2 possible solutions
- you can use an app_config.xml to limit the #concurrently running CMS tasks
Since this doesn't tell the server anything, wouldn't I still only receive CMS, no Theory, and end up with idle cores?

- you can run a separate client that runs only CMS
I find that rather unreliable. I used to do it to seperate CPU and GPU, so I can make Boinc stop using the CPU when another program is active, since Boinc doesn't have "CPU only" in their exclusive applications code for some reason. There are plenty programs which need a lot of CPU, but don't use the GPU. But the second client doesn't always start or it thinks it's the first one, or Boinc just crashes, it's quite terrible at it and too intermittant to track down.

What I do is (as I have several computers), set some to disable CMS by putting them in a seperate venue.
36) Message boards : CMS Application : no new WUs available (Message 48483)
Posted 2 Sep 2023 by Mr P Hucker
Post:
I assume CMS is given priority
Since you repeat this statement every now and then.
The server code does not have a switch to give priority to a subproject.
As long as your computer fulfils the requirements to run any subproject you get what is next in the queue.

If you feel there's such a switch, post a link to the relevant code at github.
I normally get only CMS when I ask for everything, which means CMS is always first in the queue. The other possibility is you're loading 50% of each, but a lot of folk are requesting Theory. Whatever you're doing, you're increasing CMS by giving it to those with no preference. This doesn't happen at any other project, where I assume they have a queue for each, so anyone without a preference gets some of each. I'd prefer some of each to prevent my internet connection throttling task speeds, but there's no way to do that.
37) Message boards : CMS Application : no new WUs available (Message 48481)
Posted 2 Sep 2023 by Mr P Hucker
Post:
My computers which are set to get any type of work usually get just CMS (I assume CMS is given priority on the server). But they're now getting 80% Theory, 20% CMS. I assume this was intentional? I don't mind which ones I run.
38) Message boards : CMS Application : no new WUs available (Message 48480)
Posted 2 Sep 2023 by Mr P Hucker
Post:
Indeed, I hope this means Ivan is back and well. Of all the things to attack you in a dangerous place like a collider, he goes and collides with a cabinet!
39) Questions and Answers : Unix/Linux : Is virtualbox needed for Linux and LHC? (Message 48441)
Posted 14 Aug 2023 by Mr P Hucker
Post:
You can't cd, make a new folder, paste some text into a new file and run a setup command? All just copy/paste.
Even installing CVMFS is way more complicated than that. I had to install things, then change files which didn't exist, then it simply didn't work.

Now in Windows, to install a program you double click an icon. Why is Linux still stuck in the early 90s typing commands?

Even Linux's dialog boxes don't make sense. I was editing a config file, and it was changed by the app while I was doing it. I was told this had happened, very good. But then it said "do you want to revert the file". Er... what does that mean? Revert to what? The options should have been - no changes, my changes, app's changes.
40) Questions and Answers : Unix/Linux : Is virtualbox needed for Linux and LHC? (Message 48427)
Posted 12 Aug 2023 by Mr P Hucker
Post:
https://lhcathome.cern.ch/lhcathome/forum_thread.php?id=4840#36880

That has the instructions for setting up the native apps.

Squid is something else.
Looks far too complicated for me.


Previous 20 · Next 20


©2024 CERN