Message boards : Number crunching : vboxwrapper issue
Message board moderation
Previous · 1 · 2
Author | Message |
---|---|
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 878,593 RAC: 27 ![]() ![]() |
In the slot directory you will find a floppy image file. This is attached to the VM and the user information can then be read from /dev/fd0. The vm_floppy_X.img was present where the user info couldn't be read. As I wrote it was solved by a VM-boot. With getting a new BOINC-task it wasn't an issue either. Vboxwrapper 26175 is running fine and doesn't need to rewrite vbox.log in the slot directory every minute anymore. Vbox.log in the slotdir is gone. So disk i/o's are drastically reduced. Meanwhile the Linux and Mac versions of the new vboxwrapper are available too. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 ![]() |
Vboxwrapper 26175 is running fine and doesn't need to rewrite vbox.log in the slot directory every minute anymore. Hi Crystal Pellet, Found it. How do i implement this into the cms-dev project? Do i just replace the *.exe and *.pdb wrapper files in the project folder? |
Send message Joined: 13 Feb 15 Posts: 1188 Credit: 878,593 RAC: 27 ![]() ![]() |
Vboxwrapper 26175 is running fine and doesn't need to rewrite vbox.log in the slot directory every minute anymore. You have to run it anonymously with an app_info.xml file in the project directory with the new version files together with the files mentioned in the app_info.xml and already present. Save the needed files first to a save location and reset the CMS-dev project. Stop BOINC, copy the needed files and the app_info.xml into the project-dir and restart BOINC. The contents of app_info.xml: <app_info> <app> <name>CMS</name> <user_friendly_name>CMS Simulation</user_friendly_name> </app> <file_info> <name>vboxwrapper_26175_windows_x86_64.exe</name> <executable/> </file_info> <file_info> <name>vboxwrapper_26175_windows_x86_64.pdb</name> </file_info> <file_info> <name>CMS_2015_03_13.vdi</name> </file_info> <file_info> <name>CMS_05_08_2015.xml</name> </file_info> <app_version> <app_name>CMS</app_name> <version_num>26175</version_num> <platform>windows_x86_64</platform> <avg_ncpus>1.000000</avg_ncpus> <max_ncpus>1.000000</max_ncpus> <plan_class>vbox64</plan_class> <api_version>7.5.0</api_version> <file_ref> <file_name>vboxwrapper_26175_windows_x86_64.exe</file_name> <main_program/> </file_ref> <file_ref> <file_name>vboxwrapper_26175_windows_x86_64.pdb</file_name> </file_ref> <file_ref> <file_name>CMS_2015_03_13.vdi</file_name> <open_name>vm_image.vdi</open_name> <copy_file/> </file_ref> <file_ref> <file_name>CMS_05_08_2015.xml</file_name> <open_name>vbox_job.xml</open_name> </file_ref> <dont_throttle/> <needs_network/> </app_version> </app_info> |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 ![]() |
Thanks very much, Crystal Pellet! This issue has been bugging me a lot. I will try it and let you know. |
![]() ![]() Send message Joined: 20 Jan 15 Posts: 1139 Credit: 8,310,612 RAC: 0 ![]() |
I asked Laurence to make a release with this wrapper -- see News item. vboxwrapper 26175 is now the standard version. ![]() |
![]() Send message Joined: 20 May 15 Posts: 217 Credit: 6,193,119 RAC: 0 ![]() ![]() |
I asked Laurence to make a release with this wrapper -- see News item. vboxwrapper 26175 is now the standard version. Excellent, thanks to everyone for getting this done :-) |
Send message Joined: 20 Mar 15 Posts: 14 Credit: 5,132 RAC: 0 ![]() |
In the slot directory you will find a floppy image file. This is attached to the VM and the user information can then be read from /dev/fd0. I don't know if this is still the case or not. At one time there was a bug in the readFloppy.pl script run within the guest VM where it would fail if /dev/fd0 was read-only. IIRC, it was line 58, where it attempted to open the device and perform some in-place memory modifications. I don't know what was triggering /dev/fd0 to become read-only. It happened on a previous dev box of mine. I haven't checked against my current dev box. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 ![]() |
That was implemented very quickly . I am impressed. It also appears to be working for ATLAS. Thanks everyone involved for getting this fixed. It will be interesting to see, how long it is going to take, until ATLAS is using it. |
Send message Joined: 16 Aug 15 Posts: 966 Credit: 1,211,816 RAC: 0 ![]() |
Never mind, found it. |
©2025 CERN