Posts by joeybuddy96

1) Message boards : Questions and problems : Win x64 VBox version 6.0.14 vs 6.1.12 (Message 106164)
Posted 22 Nov 2021 by Profile joeybuddy96
Post:
Just noticed the description on the downloads page says the x64 Windows bundled version of Virtual Box is 6.1.12, but the one that installs is 6.0.14. I wonder if it got rolled back to an older version or they changed their minds about using the newer version but the description didn't get updated. I can see that happening if 6.1.12 got released in the bundle but it broke stuff. I think it's less likely that they forgot to bundle 6.1.12.
2) Message boards : Projects : BOINC@TACC URL needs an update (Message 98503)
Posted 14 May 2020 by Profile joeybuddy96
Post:
BOINC@TACC recently updated their URL to use SSL. I get a notice in BOINC manager saying that I'm using "an old URL," and asks me to update to the "new URL" which is actually the old one without https in it: http://boinc.tacc.utexas.edu/. The new address ought to be "https://boinc.tacc.utexas.edu" without the quotation marks; this is the address BOINCstats/BAM! has updated to correctly. I'm pretty sure the party that needs to update the address is whoever is in charge of BOINC manager.
3) Message boards : Projects : GPU project availability (Message 98347)
Posted 6 May 2020 by Profile joeybuddy96
Post:
The BOINC wiki page on GPU crunching hasn't been updated in five years:
https://boinc.berkeley.edu/wiki/GPU_computing#Attach_to_projects_with_GPU_applications
Are there any GPU-capable projects that are missing? I know there are several that are defunct.
4) Message boards : BOINC Manager : "Postponed: vm job unmanageable restarting later" status in quchempedia (Message 97929)
Posted 21 Apr 2020 by Profile joeybuddy96
Post:
When I try to run certain quchem tasks, I get the following status sometimes: "Postponed: vm job unmanageable restarting later." I've read that it might have to do with available disk space or with the page file thing with RAM. I've got 128GB of RAM, a 64-thread CPU, an 8GB page file size, and 145GB allocated to BOINC. It's weird that I'm getting that message if that's the case. The 145GB limit in BOINC manager is in place, but the ProgramData folder has less than 90GB. I don't think page file size really has anything to do with anything aside from Windows crash reports, so I don't know where people are getting that idea from.

The quchem tasks in question have already started processing, but partway through will display the change in status from "running" to "postponed: vm job unmanageable restarting later." I could just abort the tasks, but if they're still acceptable, I don't see why I can't just wait around for them to finish. It doesn't give any conditions for when it will restart, however. If it were more detailed, e.g. "in one hour" or "once x amount of disk space is available" that would make more sense. The tasks are overdue; in the event log it says "you may not get credit for them." If they're useful though, I'm fine with letting them complete; I just don't know how to make that happen.

On a side note, given that I've got a 64-thread CPU and sometimes have more than 64 tasks taking up space on my drive due to errors or tasks that "wait to finish", what is the hypothetical maximum size of all the stuff in the ProgramData BOINC folder? It seems like the "slots" folder takes up the most space, so assume I had 64 tasks minimum, each at say, 99% completion, not counting all the other tasks that are partway completed but for whatever reason aren't currently running.




Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.