1) Message boards : ATLAS application : Atlas apparently affecting File Manager operation (Message 46751)
Posted 5 May 2022 by cuphi
Post:
Well, let's try the obvious.

$> sudo egrep -iw "warning|error|failure" /var/log/syslog


There may be lot of false positives in these results, but you have to start somewhere.
2) Message boards : ATLAS application : Atlas apparently affecting File Manager operation (Message 46749)
Posted 5 May 2022 by cuphi
Post:

When Thunar starts it may try to read directories deeper in the tree


As far as I am aware this is not the case. As I said, I don't have this issue with Thunar on openSuse. It has to be something that is unique to Linux Mint or Dark Angel's system. I don't know anything about the inner workings of CVMFS or or Singularity but I have been around Unix-like systems enough to know that deviating from the defaults is bad if you can't read the code. I am thinking that something was changed like file permissions, group membership, fstab entries, over-tuning of memory/filesystem options...it could be a wide assortment of things. Since we only have one person posting about it I really doubt it's baked into the code somewhere. I think it's much more likely to be a configuration problem. I could be wrong though. Linux Mint is a strange beast because of it's target audience.
3) Message boards : ATLAS application : Atlas apparently affecting File Manager operation (Message 46747)
Posted 5 May 2022 by cuphi
Post:
What about snapper?
4) Message boards : ATLAS application : Atlas apparently affecting File Manager operation (Message 46744)
Posted 5 May 2022 by cuphi
Post:
My openSuse boxes that run XFCE4 are not having the issue. Perhaps the LinuxMint snapshot utility is trying to capture all of CVMFS? That is really just a guess. I had issues with Mint failing tasks on my laptop so I switched it to openSuse like the rest of my network.
5) Message boards : Number crunching : Atlas Scoring algos are going insane. (Message 46694)
Posted 28 Apr 2022 by cuphi
Post:
On my laptop there is this WU: https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=187809749 : 42,648.50 84,649.43 197.45

On.my Ryzen 9 3950 there are these:

https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=187820769 : 30,082.22 58,198.01 139.27
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=187820739 : 33,903.55 65,691.93 1,569.61
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=187828931 : 34,723.24 66,261.05 5,643.76
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=187825310 : 32,956.64 62,686.52 10,081.02
6) Message boards : Number crunching : Resource Management with Systemd and Cgroups: Setting CPU Affinity in Linux (Message 46684)
Posted 26 Apr 2022 by cuphi
Post:
Because I had to read a lot of docs before I found the solution to my problem I am posting this here to help anyone else that may come along in the future.

I have a nice beefy Ryzen 9 system that I use as my crunching box and as a gaming server. It's running OpenSuse Linux. In order to keep the two roles from competing from resources I have the game servers installed in LXD containers and set them to us one half CPUs on my system. I was running my Atlas native tasks in a VM that I configured to use the other half of the CPU be Singularity can't run in an LXD container. However I was loosing a crunching power to the overhead of the VM. After stumbling through several false start I finally learned how to set CPU affinity for services using Systemd: 

sudo systemctl set-property autofs.service AllowedCPUs=8-15,24-31
sudo systemctl set-property boinc-client.service AllowedCPUs=8-15,24-31


It's as simple as that. Now I am crunching without the VM on exactly half my CPU cores while my game servers use the other half.
7) Message boards : Number crunching : Beating the Heat (Message 45158)
Posted 25 Jul 2021 by cuphi
Post:
It's very hot where I live. At least from a human body temperature perspective. I don't have a very good air conditioning unit, so to beat the heat I am stopping most of my CPU and GPU usage. Is anyone else doing the same or are you toughing it out?
8) Message boards : ATLAS application : 8 Atlas Work Units failed to download in less than 24 hours. (Message 45150)
Posted 21 Jul 2021 by cuphi
Post:
Here are the links:

https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162211
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162280
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168166322
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162256
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162352
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162517
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168162734
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=168166423
9) Message boards : Number crunching : Cannot report wus (Message 45140)
Posted 19 Jul 2021 by cuphi
Post:
Well, at least I know I am not alone in this ;-)
10) Message boards : Number crunching : How do I get the most out of my system running Atlas@home? (Message 45127)
Posted 14 Jul 2021 by cuphi
Post:
That seems to have worked. I set it to use 2 CPU's per task and it downloaded and started two more. Maybe later it will start to cache work. Thanks for the tip.
11) Message boards : Number crunching : How do I get the most out of my system running Atlas@home? (Message 45125)
Posted 14 Jul 2021 by cuphi
Post:
This is the computer in question: https://lhcathome.cern.ch/lhcathome/show_host_detail.php?hostid=10692495

I am running an AMD Ryzen 5 3600 with OpenSUSE Linux as the host OS and Rocky Linux as the guest. This is the only BOINC project I am running on this system. The host system is my squid proxy for CVMFS.

For some reason I cannot run six Atlas Native tasks at once if I want to run them single threaded. It only gets four tasks at a time and it will not even download and cache extra work.

I have plenty of free memory in the Rocky VM:

$ free -hw
                    total        used        free      shared     buffers       cache   available
Mem:           30Gi       8.7Gi        16Gi        11Mi       119Mi       5.4Gi        21Gi
Swap:         4.0Gi          0B       4.0Gi


Of the 25GB of disk space BOINC only 5.35GB is being used.
12) Questions and Answers : Unix/Linux : Is anyone testing LHC@home on Rocky Linux? (Message 45077)
Posted 22 Jun 2021 by cuphi
Post:
If you are testing LHC@home on Rocky Linux please share your experiences. I am interested in shifting my SL7 VM's to an EL release and from what I have read about it's community-based approach Rocky might be the one.



©2024 CERN