Hello, everyone.

I started a second project using asteroids@home. The computer I'm running this on has been doing seti@home projects for a couple of months. Under computing settings I have "Switch between applications every 60 minutes set." Since starting asteroids@home, no seti@home tasks have run. All of the asteroids@home tasks show "Running, high priority" under status. There are now 287 tasks in the queue for asteroids. Just today I had 100 seti@home tasks expire.

I realize that the 60 minutes setting does not mean that it will only run a task for 60 minutes. I'm wondering why asteroids@home has not hijacked BOINC and is no longer allowing seti@home to run?

Did you allocate the resource use between the two? I run four projects on a 30/30/20/20 basis. If you had a large backlog of SETI tasks and did not allocate resources when you added asteroids, I think it defaulted to 50-50 allocation. The system is now trying to "even out" computational time, hence asteroids sucking down your available time.

Anyone out there with a different read?

Under Research Share it shows:

seti@home 100(50%)
asteroids@home 100(50%)

Should I change the resource share on each to 50? I assumed that since each showed 50% I would be ok. But BOINC is refusing to let seti run at all and they're just all going past the due date and being abandoned.

Unfortunately the Replica database is offline at the moment, and that's what we need to be able to see what's been happening.
However from the information that is available there your system abandoned the WUs well before they were due. For some reason it appears it dumped all of the work in your cache on the 4th of October.

Whatever happened it has nothing to do with resource share, or deadline dates.
No issues with the date on the computer on that date?

---

Grant
Darwin NT

Asteroids at home tends to run very short deadlines, so when running in parallel with SETI does tend to run at high priority. A couple of years ago I tried running both projects on the same computer, and found that one way to prevent Asteroids dominating the computer was to significantly reduce its priority to about 10% that of SETI. At that time doing this reduced the number of Asteroids tasks being delivered to me to below that needed for the cache, instead of being substantially above what could be run before their deadlines.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

BOINC tries to allocate resources to projects as best it can but sometimes I feel it needs a bit of help. One method that can be employed is using an app_config.xml to limit the number of takes a project runs at once.

You could using a simple app_config for asteroids like this. Which would always leave one free thread open on your CPU to run SET@home.

<app_config>
   <project_max_concurrent>3</project_max_concurrent>
</app_config>

The downside to this method is that if your systems had 0 SETI@home tasks it would still only run 3 asteroids tasks.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

That's quite helpful, Hal. Thank you.

Found out that the app_config.xml must be place in that project's directory. For asteroids@home on my machine it is:
/var/lib/boinc-client/projects/asteroidsathome.net_boinc

I just restarted the boinc client and we'll see how it goes when the next task completes.

Although I'd like to get some seti@home running on this machine, asteroids@home provides much more credit. I've had a few PCs and a few PIs running seti@home since August 20th. I started asteroids@home on this one machine on October 4th and I've already earned 47,520 credits compared to having my "farm" running on seti@home and receiving 69,550 credits. Now all I need is to find the latest BOINC catalogue to see what I can buy with all of these credits. ;)

No doubt there are bugs, but one thing that a lot of folks don't realise is that BOINC resource shares are based on long term averages, not, as one might expect, "instant" averages. Based on observation the time base for the average is several weeks.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

Quite so. It's too easy to think of the BOINC project only as it applies to us individually. I also found out where the .xml file needs to go, so I think that's going to solve my issue. Not a bug at all, just having to place it properly.

Under Research Share it shows:

seti@home 100(50%)
asteroids@home 100(50%)

Should I change the resource share on each to 50? I assumed that since each showed 50% I would be ok. But BOINC is refusing to let seti run at all and they're just all going past the due date and being abandoned.

Seeing as you're treating both projects equally you should find that once your asteroids@home total credit matches your SETI credit the work should get more evenly distributed between both projects, but ATM asteroids@home is in catch up mode.

Cheers.

Ok, that makes sense, too. Thank you. I'll watch that and see how it changes when the credits match.

Just today I had 100 seti@home tasks expire.

Anyone got any theories as to why his system dumped 100 tasks on the 4th of Oct?

---

Grant
Darwin NT

Just today I had 100 seti@home tasks expire.

Anyone got any theories as to why his system dumped 100 tasks on the 4th of Oct?

- if he "Remove SETI@home project" then "Add" it again?
- if new hostid is created?
- if the same BOINC Data directory is Copied to another computer and both used at the same time?

(I think in all those cases the tasks will show as "Abandoned" (not "expired") but I don't know if anything of this happened)

---

 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 

No issues on that date.

Asteroids at home tends to run very short deadlines, so when running in parallel with SETI does tend to run at high priority. A couple of years ago I tried running both projects on the same computer, and found that one way to prevent Asteroids dominating the computer was to significantly reduce its priority to about 10% that of SETI. At that time doing this reduced the number of Asteroids tasks being delivered to me to below that needed for the cache, instead of being substantially above what could be run before their deadlines.

Thanks, Bob. I will try this.

Now that asteroids@home has surpassed seti@home (71,690 credit for Seti; 76,320 for Asteroids) and having the xml file limiting asteroids to 3 tasks at one, it still is dominating the task list. There are 282 tasks pending for asteroids (it uses only one computer) and 537 for seti (which uses 6 computers).

Under Research Share it shows:

seti@home 100(50%)
asteroids@home 100(50%)

Should I change the resource share on each to 50? I assumed that since each showed 50% I would be ok. But BOINC is refusing to let seti run at all and they're just all going past the due date and being abandoned.

Seeing as you're treating both projects equally you should find that once your asteroids@home total credit matches your SETI credit the work should get more evenly distributed between both projects, but ATM asteroids@home is in catch up mode.

Cheers.

asteroids is now ahead of seti with a credit of 76,320 vs seti at 71,690. Asteroids still dominates my task list.

I just tried what Bob Smith recommended, that is setting the asteroids resource to 10.

No luck. Asteroids is set to to have a resource share of 10 with set having 90. Asteroids continues to dominate the tasks.

The average used by BOINC is a LONG TERM average, so it will take several days for any sort of balance to be established. The first thing it has to do is process all the stupidly short deadline Asteroids tasks so it can start to work on SETI's much longer deadline tasks.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

I'll leave it and see how it goes. There are 262 asteroid tasks in the queue so it may take a while then.

Thanks.