Posts by PHILIPPE
|
21)
Message boards :
Number crunching :
Ubuntu 20.04.1 does not validate
(Message 1241)
Posted 17 Dec 2020 by PHILIPPE Post: Searching on the net , some sites speak softly about kernel differences. https://www.cnx-software.com/2019/11/25/linux-5-4-release-main-changes-arm-mips-risc-v-architectures/ https://lwn.net/Articles/791863/ https://www.omgubuntu.co.uk/2019/11/linux-5-4-kernel-release-features --------------------------------------------------------------------------------------------------------- more complete but also more technical : https://www.kernel.org/doc/html/latest/ |
|
22)
Message boards :
Number crunching :
Host ID 1388 corrupted
(Message 1232)
Posted 12 Dec 2020 by PHILIPPE Post: There is a little change in the stderr file for the last faulty computers : Before it was : <core_client_version>7.16.11</core_client_version> Now it is : <core_client_version>7.16.11</core_client_version> Segmentation fault appears.Always in line 41. Maybe here are tips to solve it (in french) |
|
23)
Message boards :
Cafe :
Serious game
(Message 1219)
Posted 27 Nov 2020 by PHILIPPE Post: Most of us know the serious game Foldit which consists in puzzles to bind proteins. I discovered that another serious game , based on the same principle exists but for rna molecules. This is eternagame If you don't know how to spend your leisure time , it can be a way , if you are interested by puzzles. |
|
24)
Message boards :
Number crunching :
Ubuntu 20.04.1 does not validate
(Message 1218)
Posted 27 Nov 2020 by PHILIPPE Post: Did you use the same way as described in your previous post to configure your installation ? The way I do dual boot is simply to use two separate disk drives (SSD), and choose which one I want in the BIOS at boot time. That keeps everything nice and separate. I don't want to know about boot loaders, whatever they are. They can only cause trouble. As for Linux, that is easy. Use the latest Ubuntu, and get the LTS (Long-Term Support version), presently 18.04.4. The intermediate versions (e.g., 19.04 or 19.10) always have strange problems logging in, among other things. They just are not tested well enough, and you will spend more time on the support forums than anything else. And when you have to choose which installation, don't use the default (first option). That does not provide a big enough root partition for BOINC (only about 20 GB as I recall). Choose "Something else". Things can get a bit tricky here, since you have to decide where you want to install BOINC, and make that directory big enough. Some people install BOINC directly in the "home" directory, which makes sense for simplifying permissions, since you are the owner of that directory and everything works (I think). In that case, you will want to make the home directory the big one, instead of the "root" directory. But I use the PPA version of BOINC (from LocutusOfBorg), which installs in the root directory. That makes installing and upgrading BOINC easy, but then you have to learn how to grant permissions and join groups, which is a big pain in Linux. You choose your poison. For a 250 GB drive, I use the following, though you have wide latitude to change it: • highlight each "free space" and click "+" and format with EXT4 • Primary partition; Mount Point "/" 200,000 MB (/dev/sda5) - EXT4 • Logical partition; Mount Point - "/boot" 512 MB (/dev/sda1) - EXT4 • Logical partition; Mount Point "home" - 5000 MB (/dev/sda6) - EXT4 • Logical partition; "EFI System" - 200 MB - FAT32 • Logical partition; Use as "swap area" - 2000 MB (/dev/sda7) - unformatted • All the rest - unformatted At least that works for me. I use Ubuntu only in dedicated crunching machines, and don't really have much in the home directory. I leave some empty space unformatted just for SSD over-provisioning purposes (to protect it), but you can use it if you want to. You will probably make enough mistakes that you have to do it over again often enough to make it easy eventually. It is just to precise to Damotbe the conditions in which the bogus occurs, more accurately. |
|
25)
Message boards :
Number crunching :
Host ID 1388 corrupted
(Message 1215)
Posted 26 Nov 2020 by PHILIPPE Post: All the computers listed (that failed) are on ubuntu 20.04. As jim1348 said , it should be good to understand why the upgrade from 18.04 to 20.04 is not successfull for this project...Time spending ,more and more computers should become unfortunately faulty. |
|
26)
Questions and Answers :
Windows :
I can no longer install Virtualbox every time...
(Message 1201)
Posted 22 Nov 2020 by PHILIPPE Post: So , it means that the initial file "VirtualBox-6.0.14-133895-Win.exe" has an error inside it. I don't see the practical solution.The weakness seems to be present only in this version of virtualbox. A user said in an another forum , that the only solution he found is to reset windows , then he has been able to install virtualbox.But it's the last solution to do if no other solution is found. maybe someone else has more skills to help you. |
|
27)
Questions and Answers :
Windows :
I can no longer install Virtualbox every time...
(Message 1199)
Posted 22 Nov 2020 by PHILIPPE Post: Before writing the command line , you have to be in the right directory. So open windows explorer , press "shift" and right click on the folder c:\download , select in the menu "open windows powershell here" , and copy the right path which is appearing. like (C:\Users\.............\Downloads) then open now a powershell windows as administrator (right click on the button start) and type the command : cd "the path pasted before" the prompt should become : C:\Users\.....\Downloads> if you type the command : ls , the name of the file downloaded should appear in the list displayed. then type your command : VirtualBox-6.0.14-133895-Win.exe --extract -path C:\VBox |
|
28)
Questions and Answers :
Windows :
I can no longer install Virtualbox every time...
(Message 1197)
Posted 22 Nov 2020 by PHILIPPE Post: Further informations on oracle site : VirtualBox 6.0.14 Silent Install (MSI) (32-bit Operating System) Software Title: Oracle VM VirtualBox 6.0.14 (32-bit) -------------------------------------------------------------------------------------------------------------------------- VirtualBox 6.0.14 Silent Install (MSI) (64-bit Operating System) Software Title: Oracle VM VirtualBox 6.0.14 (64-bit) -------------------------------------------------------------------------------------------------------------------------------- It seems that the file you don't find on the net is created by the command line : VirtualBox-6.0.14-133895-Win.exe --extract -path C:\VBox from the initial file VirtualBox-6.0.14-133895-Win.exe , you can download in oracle site. The resulting extracted file is located in C:\VBox. So can you reinstall virtualbox another time with this process and then uninstall it cleanly ? Or is it this procedure that doesn't work properly ? |
|
29)
Questions and Answers :
Windows :
I can no longer install Virtualbox every time...
(Message 1194)
Posted 21 Nov 2020 by PHILIPPE Post: sorry.double post. |
|
30)
Questions and Answers :
Windows :
I can no longer install Virtualbox every time...
(Message 1193)
Posted 21 Nov 2020 by PHILIPPE Post: Maybe try to uninstall completely VirtualBox with Revo Uninstaller. Try to install a virtualbox previous version like 5.2.32 to see if the problem is still present. It seems that Microsoft is working on a fix for this type of issue : We’re working on a fix for an issue causing some devices experiencing a KMODE_EXCEPTION bugcheck when using certain virtualization technologies. https://answers.microsoft.com/en-us/insider/forum/insider_wintp-insider_update-insiderplat_pc/cannot-update-to-build-20211-because-virtualbox/3f313552-2778-4a2f-b101-a2397db7dd17 https://answers.microsoft.com/en-us/insider/forum/insider_wintp-insider_install/keeps-asking-me-to-uninstall-virtualbox-when/6ef6e13a-047c-4fe4-bb0a-596be83ba169 Maybe better is to wait for the fix. Another post advizes : Virtualbox is installed via the WIndows MSI installer system. It appears that your Windows MSI database is bunged. It happens, it is not caused by Virtualbox, and there's things to do to fix it, using the Windows help channels. Take a look at these resources: So try the last version of the software windows repair to see if it solves your issue. Another presentation is here. |
|
31)
Questions and Answers :
Unix/Linux :
Linux vs Windows
(Message 1056)
Posted 9 Sep 2020 by PHILIPPE Post: I would not be so enthusiastic about VirtualBox.This is a great software but i learned that its behavior depends mainly of its version. I had to downgrade to 5.2.32 version to crunch this project with my computer.The next versions up to 6.1 didn't work for me.I don't know why. I prefer to stay with its old version as far as i can, to avoid to test other settings. @Tullio : If you look carefully your works units , you will see that several ones have a cpu time equal to zero. It means that there is no scientific output for these ones.This is not normal. Even if you have a very fast computer , the cpu time should be greater than 0 second (there is 3 parts in the scientific calculation). This project is very sensitive to the surrounding conditions (It is exploring the frontier between known and unknown science and there is no certitude about calculations and their result).. To maximize the reliability of your jobs , i would advise you to crunch only one work unit with VirtualBox and assign other threads to non VirtualBox projects.Then you could see if these wus with zero time cpu decrease or disappear. If it's not sufficient , then try another version of VirtualBox. If it's too much difficult to set up ,try to crunch in a Linux VM inside VirtualBox. Curiously , a linux app inside VirtualBox is more resilient and reliable than a windows app using VirtualBox for this project. Unfortunately for me, I don't have the possibilty to do it because i have a lack of ram memory. Each one has to manage with the tools at his disposition... Good luck. |
|
32)
Questions and Answers :
Unix/Linux :
Linux vs Windows
(Message 1053)
Posted 8 Sep 2020 by PHILIPPE Post: The fastest computer on this project seems to be the host ID : 2750. AuthenticAMD AMD EPYC Processor (with IBPB) Ubuntu 20.04.1 LTS 4 processors It won all the battles against its wingmen. The windows hosts can go as fast as the linux hosts but they are less reliable because of the use of VirtualBox.(it needs to be handled , cautiously , and doesn't suit with big hosts and multiple VMs running simultaneously.) Each cruncher has to do a good compromise between speed and reliability , in order to not waste his resources . Sometimes , it's better to run a bit slower but surely , especially when it is NewChem long application (several days long). Running with windows, you have to often check the wus to see if the behavior is correct.With Linux it's not necessary,the probability to fail is lower. |
|
33)
Message boards :
Number crunching :
Host ID 1388 corrupted
(Message 1043)
Posted 30 Aug 2020 by PHILIPPE Post: Just to inform : there is a boinc documentation to blacklist a host : Blacklisting hosts To blacklist a host, set its max_results_day field to -1. This will cause all of its scheduler requests to return an error message and do nothing else (no jobs or trickle messages will be sent or received). |
|
34)
Message boards :
Number crunching :
Host ID 1388 corrupted
(Message 1042)
Posted 30 Aug 2020 by PHILIPPE Post: The ID computer 1388 is corrupted. Most of the work units, it processed are "invalid (6748)"or "waiting for validation(4103)". The specific trouble is that most of its work units lasted only few seconds. It's not normal. Better would be to suspend this computer or to bannish it temporarily , untill the owner modify its settings. |
|
35)
Message boards :
Cafe :
Boinc podcast on QuChemPedIA project
(Message 1018)
Posted 15 Aug 2020 by PHILIPPE Post: Another way to speak about this project (english). This week Delta introduces us to the BOINC project QuChemPedia. https://www.boinc.network/episode/quchempedia |
|
36)
Message boards :
Number crunching :
Long work units.
(Message 987)
Posted 28 Jul 2020 by PHILIPPE Post: The best way to know if a work unit has a normal behavior, on windows hosts with VirtualBox application is to check it in the Boinc manager. You can't determine it , viewing the bar progress.(The work unit of this project are not necessarily finishable, in theory). The method :
You click , in the left panel , on the propriety button. You check in the windows pop-up , the elapsed times and the run times displayed.
|
|
37)
Questions and Answers :
Macintosh :
Computing Errors
(Message 959)
Posted 22 Jul 2020 by PHILIPPE Post: To limit the number of tasks running on your computer, you have to go to the previous link i gave you : Quchempedia-preference you change the parameter " task max" to 8 , or less if you want (even 1). At the bottom of the page. You go in boinc manager, then ,and do a read "config file" to apply the changes (item "option"). The number of task running simultaneously should change immediately. If you have a problem with the version of virtualbox , you have to try another one.(you have always the 6.1.12 version , i notice). Go to their website : older-build For some unknown reasons , some updates does'nt work, so the difficulty for the user is to choose the right one. I had the same problem with windows platform. Good luck and above all "patience".. |
|
38)
Questions and Answers :
Macintosh :
Computing Errors
(Message 955)
Posted 22 Jul 2020 by PHILIPPE Post: Looking at your stderr output for your last work unit , there is this message : 2020-07-21 21:16:46 (77678): I'm not skilled in mac-os platform , but it seems there is a trouble , with the slots number during the process. i would advise you to do another clean install of the project , but downgrading the virtualbox version to 6.1.8. instead of 6.1.12, you are using. I notice , the 6.1.8 version works fine with bozz4science.So maybe it could fix your issue. Suppress the project from the boinc manager. Unistall virtualbox 6.1.12 and install virtualbox 6.1.8 Attach the project to your boinc client. And try to limit the number of simultaneous work units to 8 at the beginning to reduce the risk of confusion between slots. If it succeeds then you will have to activate the other threads of your computer progressively and understand if you can use all the threads of your computer or not. |
|
39)
Questions and Answers :
Macintosh :
Computing Errors
(Message 951)
Posted 21 Jul 2020 by PHILIPPE Post:
|
|
40)
Message boards :
Number crunching :
missing computers
(Message 856)
Posted 8 Jun 2020 by PHILIPPE Post: Computer id : 2328 seems still active. Computer id : 2329 seems still active. Computer id : 2333 seems still active. Computer id : 2334 seems still active. Computer id : 2336 seems still active. But the fact that these computers are still active will not provide you a grant for your work. I think Damotbe should select all the wus which have been finished by only one cruncher and send these wus to another wingman. To deep learn correctly , the IA needs right datas.The fact , a wu which is only terminated by only one cruncher doesn't mean that these wu is wrong (molecule has a chance to be stable). To reward correctly , the admin should grant the cruncher which has terminated his wu when the quorum is not reached and the wu is not checked by another person , because of the admin's decision to not do a retry. It's just an opinion... |
Previous 20 · Next 20
©2024 Benoit DA MOTA - LERIA, University of Angers, France