No credit after a month?
6 Mar 2008 18:48:49 UTC
Topic 193552
(moderation:
I've been running the Einstein@Home projects in BOINC for close to a month now, yet my credit still remains at 0.00. I'm also running SETI@Home, and give equal time to the two. SETI has me at about 250 credits.
At the end of the day, I don't really care about getting credits, I'm doing this because I think it's a good idea and because I think it's a good opportunity to help advance knowledge, so the credits are only an indication that I'm not wasting my time.
Basically, that's it. I'm concerned that I'm just wasting CPU time on something that for some reason isn't working, when I could use that time on another project that I'd like to support. Since my pending credit is also at 0.00, is this a bug, or does Einstein just issue credits very, very slowly?
No credit after a month?
)
I don't know where your problem is? You have not finished any WU. First is open and second unit client detached...
My 2 Cent.
RE: I've been running the
)
Your machine was registered with the project on Feb 22 which means that it has been with the project for exactly two weeks. It received work on Feb 22 which had an expiry date of March 11. At some point (before March 06) you detached your computer and that work was abandoned. Then on March 06 ypu must have rejoined the project since you were sent more work which is still current with an expiry date of March 24. Only you can see what stage that work is at and if you don't understand the messages on the messages tab of BOINC Manager, you should cut and paste a selection of them here and ask questions. A good place to start would be the startup messages you get whenever you restart BOINC.
The project really does value your contribution.
E@H issues credits at a very similar rate per hour to what Seti does. The current science apps are pretty much debugged these days and problems with returning successful work are pretty much misconfiguration at the user end rather than server problems or science app problems. Unless you tell us in a detailed way about what is happening on your machine, it's very difficult to guess what the actual problem might be.
Cheers,
Gary.
Thanx for the replies
)
Thanx for the replies guys.
OK, I did this the day I posted(Detached a reattached). And here is my reasoning. I was looking at my account on the SETI site, and when I looked at the list of projects I was attached to it only listed itself, and a project I was attached to a while back(World Community Grid). So I thought perhaps it indicated some kind of error. So, I went to WCG, and checked my account there, but couldn't find a projects listing.
The screen saver for Einstein@Home not only lists 0 credit, but on the other side of the screen lists search information, which are both 0 as well, so I figured maybe my client was having an error, or some problem doing the work. I thought, maybe, detaching and reattaching might be the equivalnt of a system reboot when your computer starts acting screwy, if that makes sense.
As for logs, I'm not sure how the client works, so I hope this is the log you're looking for?
Does that help? If not, please let me know what file you need pasted here.
Other than everything showing up as Zero, the app seems to be working fine. The only real indication of a problem was that the credit received, and the search information are always zero.
Ah, one more thing, the "User:" up until I detached and reattached, was "None" instead of my username (LrdSothe).
Again, if you need more info, please tell me what file to get it from for you. Thanx for the help folks...
RE: Does that help? If not,
)
Please exit BOINC (BOINC Manager->File->Exit) and restart it (Start->Programs->BOINC->BOINC Manager). Open the Messages (tab or window), copy all messages. Paste them here in an answer window. That gives us some more to work with.
This does look like the
)
This does look like the "frequent restart" problem, doesn't it?
Do you have max CPU uasage set to anything below 100 % in the BOINC preferences?
CU
Bikeman
RE: As for logs, I'm not
)
Actually, no .... :-). This looks like stderr.txt from the relevant slots directory and I didn't intend for you to dig that deep :-). However since you have, and as Bikeman suggests, this looks like a particular and recent BOINC problem since you shouldn't be seeing all these restarts. In fact, if you look through the log carefully, not only is the app restarting many times after short intervals, but also there are indications that particular operations are taking much longer than they should. Here is the particular bit. Note, I've added colour (and extra space) to show each cycle clearly:-
Notes:
1. The black cycle shows that the "setting up stacks ..." never gets "done" and a restart kicks in after about 3 mins.
2. The red cycle shows the "done" comment but it took about 3.5 hours. Look at the timestamp for the following entry (06:37 -> 10.02). The setting up stacks normally takes about a minute before the "done" appears. At 10:02, the comment about opening the checkpoint is quite normal and signifies crunching starting from zero.
3. The blue cycle shows yet another restart at 10:11 so maybe a few minutes of crunching got done before this restart kicked in. However this time the "setting up stacks ..." didn't get "done" before yet another restart kicked in.
4. The green cycle is the final restart shown in your log snippet. Once again no "done" flag.
As I mentioned in my previous message, my intention was for you to use the messages tab in BOINC manager and to cut and paste from there. However, what you came up with is more indicative of the problem. This "incessant restart" behaviour is relatively new and has to be addressed by the BOINC Developers. They may have already fixed this in the latest test versions of BOINC - I don't know. As Bikeman points out, the problem seems to be triggered by setting less than 100% for your CPU usage preference. Try setting that back to 100% as a temporary workaround until the problem is fixed.
Cheers,
Gary.
Right-o, set the CPU back to
)
Right-o, set the CPU back to 100%
Also, in case it helps with the bug fixing process, I realized I'd forgotten to mention I have one of those wonderful transition processors.
Pentium 4
3.2 GHz Hyperthreading.
I also had to canabalize this box for memory, and only have half a gig in it.
I remember reading somewhere a while back that some proggies have issues with the hyperthreading part, because it can confuse them in a few ways.(Might detect as dual-core or some such, though that didn't ring true to me)
Also, I've copied out the messages, just in case that helps as well:
Thanx for all the help guys, I hope the CPU fixes the problem. And now perhaps an opinion request, is there any real reason to join a team?
RE: Right-o, set the CPU
)
Good, that should do it for now. Although I have mixed reports that it is fixed in 5.10.45, but so far the only report I had on that was for an XP 64bit. I can't guarantee that the 32bit also fixes it.
In the mean time, since I am still researching this one (and Gary, I already contacted the devs on it :-)), can you tell me what your anti virus software is?
Be part of a group you like, crunch together for greater credits. Make new friends, or enemies. If you want to continue on your own, that's perfectly all right as well.
RE: And now perhaps an
)
Oh yes, you can celebrate more milestones , yours and the ones reached by the team !
Bikeman
RE: I have one of those
)
The machine will work fine with two tasks running - one on each virtual processor. Performance will be significantly better if you don't have two tasks from the same project. Personally, I find that a 50/50 resource share with Seti with 1 E@H and 1 Seti task running simultaneously gives me good performance. In your case, with three projects, as long as any one project is not more than 50% resource share you should be OK.
That won't be an issue for BOINC.
Looks pretty normal. You can see that BOINC is seeing the 2 virtual processors and seems to be swapping your various projects between them OK. Also looks like you restarted things just after 5.00PM and have omitted some of the startup messages. Are all tasks in progress working correctly? - ie accumulating CPU time and increasing the % completed figures?
Cheers,
Gary.