| log in |
|
1)
Message boards :
Number crunching :
server issues / updates
(Message 7787)
Posted 24 Jul 2021 by Werinbert The dl/ul issues seems to start when a batch of Riesel-short tasks are loaded into the server and ends when the tasks run dry or someone kicks the server.... The kick observation is based on the server seeming to run ok after I see a "server down for maintenance" log message while crunching Riesel-short tasks. |
|
2)
Message boards :
Number crunching :
Tasks not using multiple threads
(Message 7366)
Posted 26 Feb 2021 by Werinbert Since my comment on run time, the reported run times of my tasks are now equal to the time between task sent and task received by the server.... and unfortunately not the time that the task took to run. |
|
3)
Message boards :
Number crunching :
Tasks not using multiple threads
(Message 7365)
Posted 26 Feb 2021 by Werinbert I still haven't solved the problem for the Win 10 computer. However for for the one machine that seems to be working with the multithread.....why is the run time the same (or nearly so) as the CPU time? Rather than run time around 1/4 of CPU time. |
|
4)
Message boards :
Number crunching :
Tasks not using multiple threads
(Message 7362)
Posted 25 Feb 2021 by Werinbert I don't know what is wrong with the client either... On one of my computers (Win 7 / Intel) I added back the avg_ncpus, then made the changes to the cc_config file and now things are running as expected... However, the other computer (Win 10 / Ryzen) is still not working right after making the readjustments. |
|
5)
Message boards :
Number crunching :
Tasks not using multiple threads
(Message 7359)
Posted 25 Feb 2021 by Werinbert Removed the avg_ncpus. Now Boinc allocates only one thread and the task only runs on one thread...so the result is the same as if I never set up the app_config. |
|
6)
Message boards :
Number crunching :
Tasks not using multiple threads
(Message 7357)
Posted 25 Feb 2021 by Werinbert In attempting to run "average" and "long" tasks on multiple threads I have set my app_config to <app_version> <app_name>srbase5</app_name> <cmdline>-t4</cmdline> <avg_ncpus>4</avg_ncpus> </app_version> <app_version> <app_name>srbase6</app_name> <cmdline>-t4</cmdline> <avg_ncpus>4</avg_ncpus> </app_version> And then restarted Boinc. Boinc allocates 4 threads to the task but the task only runs on one thread not 4 as desired leaving the other 3 threads running idle. |
|
7)
Message boards :
Number crunching :
Sudden credit loss
(Message 7041)
Posted 29 Nov 2020 by Werinbert From Free-DC stats last update TF -88400 (enough to change my badge from gold to silver...) |
|
8)
Message boards :
News :
Decreasing deadlines
(Message 527)
Posted 5 Jan 2015 by Werinbert I am going to agree with Steve Hawker*. We are simply looking for stationary primes not a moving target. The primes will still be there waiting in the same place whether we find them today, tomorrow, or next week. It is not like we are doing code-breaking during war time. I thought that this project is to last for a number of years. I will not change the deadline for long runners because a prime check needs much more time.??? So if a prime number takes longer to process it will be discarded for taking too long? |
|
9)
Message boards :
Number crunching :
Why 3 apps?
(Message 281)
Posted 25 Dec 2014 by Werinbert Since the run time executable that is used is llr for all three of the apps that you have made available, why are we using 3 different apps? Isn't the Sierpinski / Riesel Base app good enough to handle all of the tasks? |
|
10)
Message boards :
News :
Base S774 loaded
(Message 255)
Posted 23 Dec 2014 by Werinbert My computers are not having any high priority issues. You may want to reduce your work load buffers. When you have a lot of other tasks from other projects waiting to run Boinc goes into high priority mode for the tasks that have short deadlines. In the short term you could just (temporarily) suspend some of those idle non-SRBase tasks and Boinc will go back to normal priority. Alternatively you could set up an app_config file to limit the number of concurrent tasks. I use <app_config>
<app>
<name>srbase</name>
<max_concurrent>3</max_concurrent>
</app>
</app_config> on one of my machines to keep llr from overheating my old laptop.
|
|
11)
Message boards :
Number crunching :
Limit on tasks in progress is too small
(Message 252)
Posted 23 Dec 2014 by Werinbert Thanks. |
|
12)
Message boards :
Number crunching :
Limit on tasks in progress is too small
(Message 250)
Posted 23 Dec 2014 by Werinbert The current limit of "number of cores" is too small. As soon as a task is finished it is replaces by a task from another project as there are no tasks waiting. If you are trying to minimize the number of tasks on a computer at any given time I would say "number of cores + 2" would suffice as this would keep two tasks waiting to jump in as soon as another task finishes...and keep the other projects from jumping in instead. |
|
13)
Message boards :
News :
Base R3 loaded
(Message 234)
Posted 22 Dec 2014 by Werinbert DoctorNow wrote:
I was asking for a count not a list of primes...as in your accounts page would show something like: "You have found: 256 primes." A mere single line. Sorry if I was not clear. PG gives something similar stating: Completed tests, Credit, Primes found, and Percentage of tests resulting with a prime. |
|
14)
Message boards :
News :
Base R3 loaded
(Message 228)
Posted 21 Dec 2014 by Werinbert You should at least show the number of primes found by each user on their accounts page. |