Just noticed on one of my other computers that it is only running one task instead of the 8 it normally runs. It's an i7-3770k not sure if its doing this for the reasons stated above as well. Any insights would be appreciated.
It's very easy to do basic checks like this for yourself.
Go to your list of computers. In the far right hand column for the computer of interest, click on the last contact link to see a (possibly long) page of information which tells you what the scheduler decided to do on receiving the last request from your host. In this particular case, there is a large set of messages with some informative lines (lines 11 and 30) like:-
2020-11-20 19:37:12.1285 [PID=21149] [send] available disk 0.03 GB, work_buf_min 8640
and
2020-11-20 19:37:12.3640 [PID=21149] [send] stopping work search - insufficient disk space
In other words, you don't have enough disk space that BOINC is allowed (as set in your preferences) to use.
The GW searches (both CPU and GPU) have huge data requirements. One of my machines had about 36GB of stored data at one stage. Data files are kept for long periods simply because there may be lots more tasks to be issued that depend on the same data. It's a strategy to minimise the huge amount of bandwidth that would otherwise be required to transmit data if that data was deleted every time the task that needed it was completed.
Your GPU (GTX 970) has 4GB of RAM so should be able to handle GW GPU tasks. The above last contact log shows that your computer was attempting to get both GW GPU and CPU tasks but was denied. You do have a single GW CPU task showing in the full list of tasks as "Not started by deadline - canceled". You also have 37 GRP CPU tasks showing, all finished so none in progress.
You could potentially 'open the floodgates' by allocating more disk space to BOINC. You do that by increasing the allowances in your preferences (Computing -> Advanced -> Disk and Memory section). Before you do though, you should make sure you really have your days of work set to the values I suggested. If you look carefully at the above scheduler log you will see the work request (seconds) is still the equivalent of 28 days. Obviously, this computer hasn't seen any reduction in the days of work you had set previously.
Please set your days to 0.2 and 0.0 respectively and click on 'update' in BOINC Manager to force a communication with the project. When everything is running smoothly, you can always increase the first setting to allow you to store more tasks if you desire. Unless you really have a good reason to do otherwise, just leave the 2nd setting at 0.0.
You also need to mention how you are setting your preferences (on the website or local to the machine) and whether or not you need different preferences on different machines. Do you put machines in different 'locations' (previously known as venues) in order to give them different preferences?
As a final comment, if you can't afford to allocate a lot more disk space, you should consider disabling the GW searches and use just the GRP searches which have much, much smaller disk requirements.
Colin wrote:Just noticed on
)
It's very easy to do basic checks like this for yourself.
Go to your list of computers. In the far right hand column for the computer of interest, click on the last contact link to see a (possibly long) page of information which tells you what the scheduler decided to do on receiving the last request from your host. In this particular case, there is a large set of messages with some informative lines (lines 11 and 30) like:-
2020-11-20 19:37:12.1285 [PID=21149] [send] available disk 0.03 GB, work_buf_min 8640and
2020-11-20 19:37:12.3640 [PID=21149] [send] stopping work search - insufficient disk spaceIn other words, you don't have enough disk space that BOINC is allowed (as set in your preferences) to use.
The GW searches (both CPU and GPU) have huge data requirements. One of my machines had about 36GB of stored data at one stage. Data files are kept for long periods simply because there may be lots more tasks to be issued that depend on the same data. It's a strategy to minimise the huge amount of bandwidth that would otherwise be required to transmit data if that data was deleted every time the task that needed it was completed.
Your GPU (GTX 970) has 4GB of RAM so should be able to handle GW GPU tasks. The above last contact log shows that your computer was attempting to get both GW GPU and CPU tasks but was denied. You do have a single GW CPU task showing in the full list of tasks as "Not started by deadline - canceled". You also have 37 GRP CPU tasks showing, all finished so none in progress.
You could potentially 'open the floodgates' by allocating more disk space to BOINC. You do that by increasing the allowances in your preferences (Computing -> Advanced -> Disk and Memory section). Before you do though, you should make sure you really have your days of work set to the values I suggested. If you look carefully at the above scheduler log you will see the work request (seconds) is still the equivalent of 28 days. Obviously, this computer hasn't seen any reduction in the days of work you had set previously.
Please set your days to 0.2 and 0.0 respectively and click on 'update' in BOINC Manager to force a communication with the project. When everything is running smoothly, you can always increase the first setting to allow you to store more tasks if you desire. Unless you really have a good reason to do otherwise, just leave the 2nd setting at 0.0.
You also need to mention how you are setting your preferences (on the website or local to the machine) and whether or not you need different preferences on different machines. Do you put machines in different 'locations' (previously known as venues) in order to give them different preferences?
As a final comment, if you can't afford to allocate a lot more disk space, you should consider disabling the GW searches and use just the GRP searches which have much, much smaller disk requirements.
Cheers,
Gary.
Hello, I am facing the same
)
Hello,
I am facing the same error. Checking the logs for my laptop, I see:
Indeed, I am running Boingmgr on Manjaro ARM (Pinebook Pro is the laptop).
But checking the documentation, in theory the platform is supported by the project Einstein@Home: https://boinc.berkeley.edu/projects.php
Do I need to configure something?
I tried to search, but at the moment I haven't found solutions.
I don't see ANY ARM64
)
I don't see ANY ARM64 applications, only 32bit.
Einstein apps
Thanks for the fast
)
Thanks for the fast feedback!
I saw the list of supported OS/Architecture here: https://boinc.berkeley.edu/projects.php
But for Einstein@home is generally reported 'Linux on ARM'.
If your distro has the 32 bit
)
If your distro has the 32 bit libraries available in the 64 bit distro, you can add those libraries to run 32bit apps on your 64bit ARM device.
Then you need to change <no_alt_platform>0</no_alt_platform> in cc_config.xml to:
Then the project will send you ARM 32 applications and tasks.
Thanks also for the
)
Thanks also for the suggestion, I will check and post here updates!