Not receiving new tasks
28 Nov 2010 1:11:18 UTC
Topic 195476
(moderation:
I recently added the Einstein@Home project to my BOINC Manager while SETI@home is down, and my computer processed the first task. I hit update and sent in the results, but I have been unable to receive a new task. According to the messages:
11/27/2010 6:56:32 PM Einstein@Home Requesting new tasks
11/27/2010 6:56:33 PM Einstein@Home Scheduler request completed: got 0 new tasks
11/27/2010 6:56:33 PM Einstein@Home Message from server: No work sent
11/27/2010 6:56:33 PM Einstein@Home Message from server: Your computer has no NVIDIA GPU
11/27/2010 6:56:33 PM Einstein@Home Message from server: Arecibo Binary Pulsar Search (STSP) needs 95.37MB more disk space. You currently have 0.00 MB available and it needs 95.37 MB.
I don't understand why I was able to download/process the first task, but not subsequent ones. I have made no changes to my settings, and I have plenty of free disk space. I also had no GPU issues the first time around.
I would appreciate any help in unraveling this knot. Thank you!
Not receiving new tasks
)
Even if you have 100 GB free HDD space BOINC will not use this space if you don't permit this.
Check both:
1) The web Computing preferences
Try:
Disk and memory usage
Disk: use at most 5 GB
Disk: leave free at least 0.1 GB
Disk: use at most 20% of total
(the setting which is most restrictive will be in effect)
2) Local BOINC Preferences (BOINC Manager -> Advanced menu)
(if you use local prefs - they are "more important" (override the web Preferences))
To check how much HDD space BOINC is allowed to use:
A) Messages tab - about line #20-25 look for:
... Preferences limit disk usage to X.XX GB
B) Look in Disk tab
P.S.
I also came here (with Resource share 0) in search for backup project while SETI have no work.
I had the same HDD space "problem" (which is very easy to fix) since SETI needs (on my computer) only 100-200 MB and I used:
Disk: use at most 0.3 GB
but Einstein@Home needs 500-1000 MB (your second task is most probably another type of task (S5GC1HF, S5GC1, ABP2) which needs (temporarily) different files)
[pre] [/pre]
RE: I recently added the
)
Also because Boinc lost your video card!
http://einsteinathome.org/account/143168/computers is your pc and it clearly shows no video card. This is one of my pc's
http://einsteinathome.org/host/3454025 and you can see that is does have a video card listed. For some reason your video card drivers crashed and the easiest way to get them back is to just restart the pc. This can happen for many reason but one is that you are trying to do something else intensive, ie play games or play a movie, at the same time it is trying to crunch. If you look in the Boinc Manager, down by the clock, and do a right click on it you will see a suspend and a suspend gpu, clicking on the later should enable you to play a game or watch a movie for up to 2 hours without a problem.
Thank you for your help and
)
Thank you for your help and suggestions (both of you). I allowed for more memory usage, which helped, but did not solve, my problem. Since posting, I have rebooted several times, but to no avail. And then, yesterday, it just started working again, processing data as though nothing were wrong. So, I don't know what happened, but it's working again, and I appreciate your help in this matter.
RE: I allowed for more
)
Maybe because it wasn't your problem in the first place? ;-)
The error message was about lack of disk space, not main memory.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
I am receiving a message that
)
I am receiving a message that I have "reached daily quota of 2 tasks." What does it mean and how can I get the 10 days of work that this computer was once getting?
RE: ... What does it mean
)
It means that you have already received your allowed 2 tasks per core for the current day. Normally your allowance is 32 tasks per core per day. If your host is suffering computation errors (as yours is) your allowance will reduce by one for every failed task. You need to investigate why tasks are failing on that machine. Once you solve that, each successfully completed task will cause your depleted quota to double, so you would soon get back to 32/day.
You should click on the taskIDs of the failed tasks on the website so that you can review the error messages sent back with the failed tasks to gain some idea of what might be wrong. For the messages I've perused, I suspect you have some sort of hardware issue, possibly flakey RAM or perhaps related to excess heat and/or overclocking. I say that because the two tasks that failed initially were a substantial way though crunching when the error occurred. One of the error messages was about "too many exit(0)s" and the other was about a "missing or corrupt data file". The second message is probably bogus. The task would not have started crunching in the first place if the message were true. However I have seen exactly this sort of message many times, when the initial crash is triggered by hardware/heat/overclocking. Almost invariably, the data file is neither missing nor corrupt. It's pretty easy to find that out by checking the MD5 checksum (provided BOINC hasn't already removed the supposedly corrupt file).
Return some good tasks ;-).
Why do you feel you need a 10 day cache? With only a 14 day deadline and a reliable, high availability project, you don't really need that much.
Cheers,
Gary.