I have six machines of varying ages crunching on a variety of projects.

Two machines have recently begun to accomlish absolutely nothing. Both are modern machines, one an i7-840 laptop with 8 gigs of DRAM and crunching on 5 threads with 7.0.28 on a variety of projects. The other is an i7-920 desktop with 6 gigs of DRAM and crunching 7.0.28 on 5 threads. Both machines have been cleaned and vacuumed, both are running with tthrottle because of excessive heat. Both are running the 5 threads at 100%.

Two machines, an i7-3930k with 16 gigs of DRAM is just fine on 7.0.28, and an i5-520M with 4 gigs of DRAM is fine on 6.12.34, both doing plenty of work. Both are also using tthrottle.

When I started to see problems about two weeks ago, I un-installed 6.12.34, deleted BOINC and BOINC Data, edited the Registry to eliminate BOINC and WCG, re-installed with 7.0.28, copied in appropriate cc_config files mostly to ignore GPU's. I re-attached to the various projects, and the problems have continued.

On both machines, the Remaining time is moving up not down.

I do not kow what to try next. Any help will be appreciated.

---

http://sciencesprings.wordpress.com
http://facebook.com/sciencesprings

Mind telling what exactly you added in the cc_config.xml file?
Mind telling which projects you're talking about?
Or which applications?
Have you taken into account that since you cleaned up everything BOINC, that BOINC 7 is really starting off clean and that it now takes time for it to learn all about the applications again?
Are you running 24/7, or other?
With which cache, as in what are your settings for Maintain enough tasks to keep busy for at least X days and for ... and up to an additional X days?
Where did you set these preferences?

---

Mind telling what exactly you added in the cc_config.xml file?
Mind telling which projects you're talking about?
Or which applications?
Have you taken into account that since you cleaned up everything BOINC, that BOINC 7 is really starting off clean and that it now takes time for it to learn all about the applications again?
Are you running 24/7, or other?
With which cache, as in what are your settings for Maintain enough tasks to keep busy for at least X days and for ... and up to an additional X days?
Where did you set these preferences?

Ach, as usual, a good friend looking to rescue the hapless soul.

Here is the cc_config file, added into B OINC Data just to eliminate GPU's:

<cc_config>
<log_flags>
<sched_op_debug>1</sched_op_debug>
<debt_debug>0</debt_debug>
<work_fetch_debug>0</work_fetch_debug>
</log_flags>
<options>
<client_version_check_url>http://www.worldcommunitygrid.org/download.php?xml=1</client_version_check_url>
<client_download_url>http://www.worldcommunitygrid.org/download.php</client_download_url>
<network_test_url>http://www.ibm.com/</network_test_url>
<start_delay>120</start_delay>
<no_gpus>1</no_gpus>
</options>
</cc_config>

Projects
Asteroids, CAS, Einstein, eOn2, LHC@home, Milky Way, SAT@home, Cosmology@home, WCG.

Running 24/7

Caches: "Minimum work buffer" (used to read "Connect every...") I leave at the default. the i7-840 installed this as "0.00 days", the i7-920 installed this as "0.10 days". The "Max additional .." on both is 1.25.

The preferences are set in BOINC Manager\Tools\BOINC- Preferences on each computer.

---

http://sciencesprings.wordpress.com
http://facebook.com/sciencesprings

OK, so what do you then mean by 'nothing'?
Else, what science applications do this explicitly?
You have taken into account, etc.?

Have you asked at the projects whose applications do this if this is something they know of, or that others see this as well?

---

OK, so what do you then mean by 'nothing'?
Else, what science applications do this explicitly?
You have taken into account, etc.?

Have you asked at the projects whose applications do this if this is something they know of, or that others see this as well?

I mean, in Boinc Manager\Statisics they both are flat lined at zero on host total. User total figures are correct on the i7-840. BUt in the i7-920, the only project which shows anything in User total is LHC@home. So, this is definitely whacky.

I have not gone to the projects because all of the projects are running not only on these two machines but two others, and they are both fine. Also, the Statistics show up correctly for both User total and Host total. So, I believe that the problem(s) are mine and not the projects'. Boohoo.

You know, I think that it is stuff like this which has some number of crunchers just leave the field. I mean, I know, lots of folks leave for economic, family, health, work, and school issues. But some of this stuff just defies description. I have never been a quitter. I have always brought my problems to the forums, because I want stuff out in the open. Lots of time I am correct, lots of time I get corrected, and that is all fine. I just think that maybe others might be helped when they come looking for situations similar to theirs.

---

http://sciencesprings.wordpress.com
http://facebook.com/sciencesprings

I mean, in Boinc Manager\Statisics they both are flat lined at zero on host total.

Don't take this too personal but you seem to have a habit of assuming too much on far too little evidence. Or maybe it's just something about the way you express yourself. You see when you say words like "they are accomplishing nothing" what that suggests to most readers here is that the machines aren't crunching any tasks. It means they aren't receiving any tasks, aren't crunching anything, aren't returning any results. You can get advice quicker if you use words that have unambiguous meaning. Phrases like "isn't receiving tasks" or "has tasks listed on Tasks tab in the manager but they're not making any progress" or "tasks are crunching but all ending with Compute Error"... those are words we understand. "Accomplishing nothing" could mean lots of things.

OK, so in this case you're telling us they've flat lined. Again, we don't have a clue what "flat lined" means to you. Then it turns out you're looking at the graphs on the Statistics tab..... finally we are getting somewhere!!! Finally the images that you see in your brain are traversing the distance between us and coalescing in our brains. Why? Because finally you used unambiguous words, words that we all know the meaning of.

See you speak of the frustration people often feel when running BOINC and kind of hint at the difficulty in coming here and getting advice. Have you ever, figuratively speaking, walked around to the other side of the connection that joins our machines and sat yourself behind the keyboard of those who attempt to give help here? Trust me... there is frustration and difficulty on both sides of the table. It is very difficult sometimes for us to understand what's going on on your host. If you want quick and easy help then use language that describes the problem in common terms. If you had said right off the top that the graphs on the stats pages in BOINC manager are flat I know for sure all of us on this end would have felt a lot less like "oh crap, the vagueness begins again" and would have told you what I am about to tell you....

The graphs are derived from data about the tasks. If the data is not present or something goes wrong with the derivation process then the graphs will not be an accurate reflection of what's actually happening re: crunching. And this is an example of where I say you assume too much. You assume that just because the graphs are flat there is no crunching going on. Or at least that seems to be what you are saying when you say "accomplishing nothing".

If that is what you are trying to say... that no tasks are being crunched.... then there are several other far more reliable indicators you should be looking at to check see if what the graphs say is actually true. The graphs are the least reliable indicator. You should be looking at the Tasks tab in Advanced View to see if tasks are coming, crunching from 0% up to 100% and then uploading. You should be looking at your Lists of Tasks on project websites to see if your hosts are returning results that validate. You should be looking at your List of Hosts at each project to see if your hosts are at least contacting your projects. If things don't appear to be kosher then you should be looking in BOINC manager's Event Log to see if the host is issuing scheduler requests and to see what the outcome of those requests are.

But to hear you say you've looked at the graphs in BOINC manager and have concluded that those hosts are accomplishing nothing is just so bizarre to me I can hardly believe it. But please... don't think that means I don't like you or that I think you're stupid, it does not. I want to help. But I don't know how when you post what you post. Tough words, don't let them get you down. My feeling truly is that I want to help not hurt.

I often get the idea that for many of those who come seeking help we help givers should be using software that allows us to login on your machine and take control of it for a few minutes and see exactly what you see on your screen. Like the way your mechanic takes your car for a test drive when you tell him "there's a weird noise when I get up to 60 MPH and I don't have a clue what it is". We would control the mouse cursor and click on things with our mouse and type stuff on your computer from our keyboard. You could watch what we do. We could show you exactly what you need to click and what to look at. Coupled with a Skype connection or some other kind of voice connection it could be a very powerful tool for teaching and learning. Interested in trying something like that?

OK, so what do you then mean by 'nothing'?
Else, what science applications do this explicitly?
You have taken into account, etc.?

Have you asked at the projects whose applications do this if this is something they know of, or that others see this as well?

I mean, in Boinc Manager\Statisics they both are flat lined at zero on host total. User total figures are correct on the i7-840. BUt in the i7-920, the only project which shows anything in User total is LHC@home. So, this is definitely whacky.

I have not gone to the projects because all of the projects are running not only on these two machines but two others, and they are both fine. Also, the Statistics show up correctly for both User total and Host total. So, I believe that the problem(s) are mine and not the projects'. Boohoo.

Are you using the advanced view? If you go the the Tasks view what do you see? Maybe all the tasks are having computation errors because of your heat problems?

I'm on the Seti project with BOINC. I'm downloading files with some files not downloading (halting), but others do finish the download. My problem is that the downloaded files are NOT becoming tasks for my quad core to process. It was doing a fine job of that, using all 8 processors yesterday. Now, it is processing nothing.

I'm telling the software to use the CPU and GPU all the time.

So, as the new kid, what button do I push to get the new files to process?

---

I guess I was impatient. For reasons known only to the program, it suddenly started processing tasks again. I have no idea why. I didn't do anything to encourage it, other than posting my question.

Unless it chokes again, I'll consider the problem solved.

---