Hi BOINCers,

Hi Guys,
I run MW@H and SETI 50/50 but since downloading 7.0.25 all my MW@H wus NBody and PS now seem to be running at high priority.
Guys I have a two CPU core Intel E4700 Acer Aspire M1640 using Vista HP sp2 with no GPU worth mentioning. What is now happening is that SETI is being effectively excluded by MW@H which is running wus with a report date 13 days hence in high priority.
I am not confident about "applying patches" or any other such fix as recommended on the MW@H Number Crunching message board so I have been suspending all but one MW@H task to enable SETI to run. This means that I have to set my alarm to go off each time the current MW@H task is due to complete so that I can enable the next task. THIS IS NOT A VIABLE OPTION in the long term.

CAN SOMEBODY PLEASE advise me how to handle the glitch in the 7.0.25 BOINC Manager so that normal working can be resumed. Can I still get the previous version of BOINC (6 something) which did not do this?

Thanks
John

7.0.27 has a fix for the "client: fix bug that caused a project's jobs to all be run EDF if the project has the <dont_use_dcf> flag set." problem.

In the mean time, if you want to return to 6.12, follow the following steps:

Returning to BOINC 6
Due to changes to how BOINC 7 writes details about work to the client_state.xml file, you can't easily return to BOINC 6 when BOINC 7 has run on your system. Not without you having made an earlier backup of the data directory, and even then it'll be troublesome. Any work you have gotten since you ran 7.0, will be 'lost' when you return to 6. Only projects that have "resend lost tasks" on will resend them to you, but not all projects do this due to the severe strain on the database. (Only Seti and Einstein have it on, I think).

So here's what you can do to return to BOINC 6 in an easier manner:
- Set No new tasks on all your projects, run all the work to completion, upload & report it.
- Uninstall BOINC.
- Go to the data directory and remove the client_state.xml and client_state_prev.xml files.
- Install BOINC 6.xx, whichever flavor you liked.
- Run that.
(- Set Allow new tasks on all your projects.)

That's all.

---

Hi Ageless/Jord,

thanks for your reply. I think that going back to 6.xx is in more sense than one a retrograde step so I probably wont do that, at least not without further thought.

I am managing, not without a lot of hassle, to continue crunching using the suspend/resume tactic. I may just keep on doing that because I have, for me. a large cache of tasks to calculate

You say that 7.0.27 has the fix do you have any estimate of the time when that will be released?

Thanks for the help and advice.

John

You say that 7.0.27 has the fix do you have any estimate of the time when that will be released?

No, last time I said that a version was imminent for the same day development hit several snags and had to postpone for several days. So I won't dare to guess anymore. When there's enough bugs found & fixed, I think. ;)

---

You say that 7.0.27 has the fix do you have any estimate of the time when that will be released?

No, last time I said that a version was imminent for the same day development hit several snags and had to postpone for several days. So I won't dare to guess anymore. When there's enough bugs found & fixed, I think. ;)

I am waiting for .27 also. The same reason. :)

Yeehoo,

I just installed 7.0.27 and no more high priority running. I have not tested it fully but so far so good!

John

Hi,

to update everyone the move to 7.0.27 caused the removal of the "high priority" tag on wus. However it is taking some time before my pc is running 50/50 SETI and MW@H without my interference.

I will leave it to sort itself out but it appears to be taking some time. If I didn't think that I knew better I would think that MW@H is still running at high priority but the tag has been removed.

Has anyone else noticed this? If so please let me know.

John

On the 3rd May I downloaded 7.0.27 and the high priority tag disappeared form MW@H. Unfortunately the problem with MW@H still running SETI off my system persists. Unless I intervene by suspending outstanding MW@H tasks and allowing SETI to run then SETI will not run.

I am unsure whether this is a BOINC or MW@H problem so I have posted here and on the MW@H number crunching forum. But although the high priority tag has been removed I still have the same problem using 7.0.27 as I had using 7.0.25.(see my earlier posts on this thread)

Is there another way to sort this should I change something in the project preferences?

Thanks for any suggestions.
John

Hi BOINCers,

I am now running 7.0.28 and MW@H is still hogging all the CPU time. Unless I intervene by suspending MW@H tasks waiting to run then they run to the exclusion of my other project SETI. Although the "high priority" tag was removed when I went to 7.0.27 the MW@H tasks still push the SETI tasks off the CPU.

This is a long-standing problem for me (see my earlier posts) that I cannot seem to solve so I would be happy if someone could suggest anything to try.

Thanks
John

What are your cache settings?

Claggy

Hi Claggy,

I have looked at my SETI preferences and I maintain enough tasks to keep busy for 2 days and up to an additional 7 days.MW@H has resource share set to 100.
SETI has been set to use at most 100% of CPU time. I hope that this is what you mean by cache settings if not then please advise.

On BOINC manager Projects both resource shares are set to 100 (50.00%)

Thanks for your reply

John

If you look at the top of the screen where you set that SETI '2 days plus 7 days' preference, it says:

These apply to all BOINC projects in which you participate.
On computers participating in multiple projects, the most recently modified preferences will be used.

So, you'll have set the same preferences for Milkyway as well.

Milkyway seems to have a turnround period - from issue to deadline - of 12 days. With your settings, you're asking Milkyway to keep 9 days' worth of tasks on hand, and finish them no later than 10 days after issue.

That's because "maintain enough tasks" is designed to cover the case where you can't connect to the internet for - in your case - two days. There's no point in finishing a task 30 seconds before deadline if you're not going to be able to report it until the next connection, in two days time. So 'maintain enough' doubles up as 'and finish them this early, too'.

Reduce the size of the cache, and Milkyway will relax a bit.

Hi Richard,

thanks for the quick reply. Setting the preferences for MW@H is different from SETI in that there does not seem to be the 2 days plus 7 days. MW@H only sends 6 tasks at a time and I can calculate them well within the two days. Instead of running 1 core MW@H and one on SETI both do MW@H and then request another 6.

I have pasted the settings for MW@H below can you advise me which one needs to be changed. Can you also say why in BOINC Manager Projects Resource share for both is shown as 100(50%)?

Resource share
Determines the proportion of your computer's resources allocated to this project. Example: if you participate in two BOINC projects with resource shares of 100 and 200, the first will get 1/3 of your resources and the second will get 2/3.
Use CPU
Enforced by version 6.10+
Use AMD (ATI) GPU
Enforced by version 6.10+
Use NVIDIA GPU
Enforced by version 6.10+
Is it OK for MilkyWay@Home and your team (if any) to email you?
Emails will be sent from boinc@cs.rpi.edu; make sure your spam filter accepts this address.
Should MilkyWay@Home show your computers on its web site?
Default computer location
Maximum CPU % for graphics
0 ... 100
Run only the selected applications MilkyWay@Home
MilkyWay@Home N-Body Simulation
If no work for selected applications is available, accept work from other applications?
Frequency (in Hz) that should try to complete individual work chunks. Higher numbers may run slower but will provide a more responsive system. Lower may be faster but more laggy.
default 60 (corresponds to 60 fps)
Period (in milliseconds) to check for scene updates
Change how fast the view randomly moves around
Make particles bigger or smaller
Make particles bigger or smaller when using uglier particles

Thanks for the help.
John

p.s. are you saying that if I cut SETI preferences to 2 days plus 2 days then that will help MW@H run at 50%?

You're looking in two different places.

Both project account pages have three sets of preferences:

* When and how BOINC uses your computer
* Message boards and private messages
* Preferences for this project

At SETI, you're looking at the top set, which includes the 'enough tasks' and 'additional tasks' values - which are transmitted from project to project, so the same for all BOINC projects.

At Milkyway, you're looking at the bottom set, which is a different collection of numbers entirely - as your paste shows.

Hi Richard,

MW@H then uses the 2 days plus 7 that I set for SETI. What I was doing is going to SETI and selecting preferences and then going to MW@H and doing the same. SETI was my first project so I suppose that it is sort of logical, although to me confusing, that it has this extra, cross project setting ability

What I have now done is to go to SETI and change the preferences to 2 days plus 2. As I have just changed them I will need to wait until the system settles down again to see if that does the trick.

Does all this make sense to you? Have I done the right thing?

Thank you very much for the help, I certainly need it! I will post again her if after a while it turns out to be the answer to this problem.

John

Computing preferences is a Global preference, you only need to set it at one project, once Boinc contacts that project it'll update it's global preferences file,
Once Boinc contacts all your other projects, it'll update them with the latest preferences,

Edit: typo's edited out, Thanks.

Claggy

Computing preferences is a Global preference, you only need to set it one project, once Boinc contacts that project it'll update it's global preferences file,
Once Boinc contacts all your other projects, it'll update them with the fastest preferences,

Claggy

That would be 'latest' preferences - the preferences most recently set, no matter where you set them.

Thanks Claggy and Richard,

I have modified my preferences so I will let you know here if things eventually balance out.

John

Hi Claggy, Richard, Ageless and others who posted to this theme,

I am afraid that I am still where I was when I posted on this last April. I am still running 7.0.28 with the 2+2 preferences but unless I restrict MW@H using the suspend and resume tactic it will kick SETI off and only run MW@H.

I thought that if I gave it a few months to settle down then I would get back to 50/50 on both projects but this has not happened. Although the "high priority" tag for MW@H disappeared after 7.0.25 it is , in effect, behaving as if it was still there.

Does anybody have any ideas? If so, thanks for posting them.

John

Hi Claggy, Richard, Ageless and others who posted to this theme,

I am afraid that I am still where I was when I posted on this last April. I am still running 7.0.28 with the 2+2 preferences but unless I restrict MW@H using the suspend and resume tactic it will kick SETI off and only run MW@H.

I thought that if I gave it a few months to settle down then I would get back to 50/50 on both projects but this has not happened. Although the "high priority" tag for MW@H disappeared after 7.0.25 it is , in effect, behaving as if it was still there.

Does anybody have any ideas? If so, thanks for posting them.

John

Hey there.
I had written up some strange-looking observations about the n-Body-108
There is a MW forum that asked for input about their beta-program. [edit-clarify]

see
http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3171

It explained a lot to me.
I run Linux. (may not matter)
I had set my BOINC preferences ( cpu % cores utilized) to 50 % and every other time a certain WU was downloaded the actual utilization went from
62.5% to 100% to 62.5% to 100% , etc.....

I *think* this relates to your observations.

Jay