Hi all !

Its been 2 days now and my BOINC is not getting any more work units...

Why ?

Check your BOINC settings. There is work available.

But one of your machines has only 2 GB of memory. That is not enough. I wouldn't try with less than 8 GB.

---

I have 30 Gigs of RAM.

So why do I have no GPUGRID WUs. And only GPUGRID work I have set to only.
And to check settings for what. Work was just fine for me until it wasn't. What does having to check settings have to do with anything Jim?

Rosetta still has "14943" WU ready to send and no work?

4/14/2020 9:12:45 PM | Rosetta@home | No tasks sent
4/14/2020 9:12:45 PM | Rosetta@home | Tasks won't finish in time: BOINC runs 97.7% of the time; computation is enabled 99.6% of that
4/14/2020 9:12:45 PM | Rosetta@home | Project requested delay of 7 seconds
4/14/2020 9:24:05 PM | Rosetta@home | No tasks sent
4/14/2020 9:24:05 PM | Rosetta@home | Tasks won't finish in time: BOINC runs 97.7% of the time; computation is enabled 99.6% of that
4/14/2020 9:24:05 PM | Rosetta@home | Project requested delay of 7 seconds
4/14/2020 9:47:14 PM | Rosetta@home | No tasks sent
4/14/2020 9:47:14 PM | Rosetta@home | Tasks won't finish in time: BOINC runs 97.7% of the time; computation is enabled 99.6% of that
4/14/2020 9:47:14 PM | Rosetta@home | Project requested delay of 7 seconds
4/14/2020 10:13:19 PM | Rosetta@home | No tasks sent
4/14/2020 10:13:19 PM | Rosetta@home | Tasks won't finish in time: BOINC runs 97.7% of the time; computation is enabled 99.6% of that
4/14/2020 10:13:19 PM | Rosetta@home | Project requested delay of 7 seconds

You should set BOINC to run 100% of the time.
It is the "Use at most 100% of the CPU time" setting.

Also, do not suspend when computer is in use.

---

Your average turnaround time is 4.8 days, but the WUs currently being sent have 3 day deadlines. If you are unable to find a setting that improves the expected runtime, you might reset the project. But I'm not certain if that will reset those statistics.

Looking a the resource shares with your other projects, what % of the system is the resource share for R@h?

---

Rosetta Moderator: Mod.Sense

Well... that is a dedicated machine for BOINC !

It never had any problems with Seti@home ... :-

I wanted to help with that Covid19 thing ... :-(




And please others : My thread ! Create your own ! ;-)

You're right, we got sidetracked there. Your hosts are returning work in about a day, so that is good.

One of your hosts is 2 CPUs with 2GB of memory. I would suggest that you set that one to use 50% of the CPUs.

When you update to the project, there should be some messages in the event log. Firstly confirm if the machine is requesting new CPU tasks, and then what reply does it get? Also, are you having similar issue on both machines? Or, which one?

And please others : My thread ! Create your own ! ;-)

It's better to use an existing thread than start a new one. We'd rather answer a question once, not 100 times.
Your systems need to return the work they have, before they will get more.
One system has returned some work, the other none.

---

Grant
Darwin NT

I am having the same issue where Rosetta is not giving me more work. I have it set up so there is a limit of 3 tasks at a time for Rosetta and 3 tasks for Einstein. It seems that Einstein is taking all the room for tasks so Rosetta won't send more. I am not sure how I fix this issue without just suspending Einstein until Rosetta gives me more work.

Examine how much disk space BOINC is allowed to use. And then I believe that is divided between your projects based on resource shares. R@h should have about 10GB of disk to have enough room to run 3 and have some in the cache of ready work.

---

Rosetta Moderator: Mod.Sense

Disk space doesn't seem to be the issue, as I have 50 gigs that BOINC has available to it. They are both on equal resources sharing, but for some reason I can't seem to get Rosetta tasks.

The only host registered to the account you are posting from has a couple dozen active work units.

---

Rosetta Moderator: Mod.Sense

It has some work now, but it was all manually gotten work. I will keep an eye to see if it continues to grab more on its own.

I am having the same issue where Rosetta is not giving me more work. I have it set up so there is a limit of 3 tasks at a time for Rosetta and 3 tasks for Einstein.

BOINC doesn't work that way, and it definitely can't work that way due to the differences between the projects.
Einstein Tasks run until they are finished, the more powerful your system, the sooner they will finish. Rosetta Tasks run for a set period of time.

What runs & when will be determined by your Resource share settings. But until the Estimated completion times for Rosetta start to match reality it will take a while for the Manage to figure out what is going on & balance things out. The larger your cache, and the more projects you do, the longer it will take for thigs to settle down.

Reduce the size of you cache, and it will help things to sort themselves out sooner.

Other	
                                Store at least 0.3 days of work
                     Store up to an additional 0.02 days of work

---

Grant
Darwin NT

I am going to exclusively run Rosetta with just 3 tasks at a time and run nothing else for a while until it can get a better sense of how much work it can do. Then I will try adding other projects in to see if they will play nice together.

I am going to exclusively run Rosetta with just 3 tasks at a time and run nothing else for a while until it can get a better sense of how much work it can do. Then I will try adding other projects in to see if they will play nice together.

They won't at first, because for each project it has to figure out how much time it takes to complete a Task, and that takes a minimum of 10 Valid Tasks. And if a project has more than 1 application, then for each application is has to figure out the time for them as well. And then you add another project, and it will have to figure out the times for those applications.
And the entire time it will be trying to honour your Resource share settings, so it will do lots of one project, then lots of the other project, and maybe oscillate back and forth between them a bit for a while as it gets the Estimated Completion times sorted out. Then when you add another project, it has to sort out the new balance as Estimated completion times for that project settle down. Add another project, and it has to do it all over again.
And then varying the number of cores & threads available to BOINC will add yet another factor in increasing the time it takes for things to stabilise.

So a very small cache, set the number of Cores/Threads BOINC can use, select the projects & the tasks you want it to do (for each project you may need to set specific settings there for certain type of work- to do or not to do) set your Resource shares, and then just let it be for a few weeks while it gets thing sorted out. The smaller the cache, the sooner it can sort things out. The more things you fiddle, the longer it will take to sort them out.

---

Grant
Darwin NT