Setting BOINC disk space on headless Ubuntu 18.04

supdood

Well-Known Member
USA team member
Hi all. I recently got the following error message on one of my free Google Cloud VMs (e2-micro).

Message from server: Mapping Cancer Markers needs 46.04MB more disk space. You currently have 453.96 MB available and it needs 500.00 MB.

Usually the VM can run two tasks concurrently, but since I've been getting this error it only downloads and runs a single task at a time. Reviewing the overall disk space on the VM, there is about 2 GB free. I'm familiar with setting the BOINC disk space allocation using the Windows and Linux BOINC Manager GUIs, but not sure how to do so from the command line. Reviewing the boinccmd documentation did not provide any leads.

Any suggestions would be much appreciated!
 

supdood

Well-Known Member
USA team member
Thanks, Ronnie! This lead me directly to the solution. It has been so long since I had to mess with project-specific device profiles other than to change which sub-project I wanted that I forgot that disk usage limits were part of the global preferences sent from the project profile. I had gotten used to setting the overrides via the BOINC Manager and somewhere along the way lost the fact that it was an override and not the only way to set the value.

I simply updated the values for the specific device profile in WCG and then issues an update command. On the following update the tasks flowed in and we are back to 100% CPU utilization.
 

RONNIE

Well-Known Member
USA team member
Thanks, Ronnie! This lead me directly to the solution. It has been so long since I had to mess with project-specific device profiles other than to change which sub-project I wanted that I forgot that disk usage limits were part of the global preferences sent from the project profile. I had gotten used to setting the overrides via the BOINC Manager and somewhere along the way lost the fact that it was an override and not the only way to set the value.

I simply updated the values for the specific device profile in WCG and then issues an update command. On the following update the tasks flowed in and we are back to 100% CPU utilization.
Yes, the solution you've adopted is easier. Great!
 
Top