I'm running my first WU now. The program seems to be running properly except that the progress percent is not changing and the time to completion is not dropping. The WU is about 10 minutes in now with no change.


FYI: 900 Mhtz AMD Thunderbird
768 MB RAM
Windows XP Pro w/ SP2
BOINC v.4.45


Update: Restarted BOINC after 15 minutes. The WU seems to have started over now, progress and time to completion still locked.

---

In wildness is the preservation of the world, so seek the Wolf in thy self...

I'm running my first WU now. The program seems to be running properly except that the progress percent is not changing and the time to completion is not dropping.

I have experienced the same. I think there's something wrong with the time estimation too. At first, I was told the WU could not be completed in time. I then suspended all other projects on this machine and it got started.

After 4:45 hours of crunching, the WU shows as 66.67% complete (P4/2.8) and has been stuck on that for some time. I didn't see it get there as I went out, so don't know how often it was updated (every 33.33% perhaps?)

Mobile Pentium 4 / 2.8GHz No HT
256MB RAM
Windows XP Pro
BOINC 4.45

It also says there's still 65 hours to go for the remaining 33%.

---

*** Join BOINC@Australia today ***

My first WU shows an estimated completion time of 156 hours. Is this what you get when you combine Protein Predictor with CPDN?

---

[email protected]

I carefully watched the first few WU's that my machines crunched and observed that they often appear to get stuck at around 66%.
They get past it eventually so don't worry.

The long estimated finished time is a bit of a problem though, it put boinc straight into shortest deadline mode as it thought my machines were overcommitted.

---

I carefully watched the first few WU's that my machines crunched and observed that they often appear to get stuck at around 66%.
They get past it eventually so don't worry.

The long estimated finished time is a bit of a problem though, it put boinc straight into shortest deadline mode as it thought my machines were overcommitted.

There are two parts to the workunits that are currently being processed. The first is the prediction of low-resolution structures (the first 66%, the next batch will be the first 83%). The second is high-resolution refinement (all atoms are considered), and this takes a bit more time. Since the percentages are currently based on the number of structures (not factoring in the different methods), they are not perfectly accurate. That is why you are seeing a lag after 66% as this is when the refinement step starts. The types of work units including the time they take to finish will change with time as we do various types of experiments.

The types of work units including the time they take to finish will change with time as we do various types of experiments.

Thanks for the explanation. However, the estimated times are so far out that it causes BOINC major headaches. I have a WU running on a 2.8GHz Pentium 4 and BOINC complained from the word go that it would not be able to complete the WU before the deadline (estimating it would take over a week) and would not start on it. I could only get it to start by suspending all other projects.

I know the estimates will never be accurate, but 150+ hours estimated for a job that might take 12-18 hours (judging by results on other computers) is not even close. If this can't be sorted out you may end up with a lot of thrashed work units (and very few volunteers) if BOINC refuses to crunch.

I cancelled my first WU after 15 minutes because it seemed stuck in an endless loop (0.00%). Others will likely do the same?

Maybe a good idea to change the % done to some number right after the start, until a more accurate way is found. This way the user is more likely to get the impression something is happening. Predictor work units display 13% done after a few seconds, and then gradually start counting.

Also, I'd suggest you put on the home page the project is in Beta (unless I overlooked). People who are afraid that the testing might interfere with other projects can then decide to stay away for a while.

---

BOINC.BE: For Belgians who love the smell of glowing red cpu's in the morning
Tutta55's Lair

Yes, and those of us who write documentation like to get it right! :)

If one or more of the project people want to become an editor on the "Unofficial" Wiki, send me an e-mail ([email protected]) and I will add you to the editors list, be sure to identify that you are a project developer for this project. We do like to have people that work on the projects so that they can link their project in with all the other projects and the rest of the BOINC documentation.

You can look at our "class of the world" project with CPDN where Chris R. is doing supurb work ... tough standard to live up too ...

Oh, and you can add a link to the Wiki if you would like ...

http://boinc-doc.net/boinc-wiki/

---

I knew this wouldn't stay a wiki free zone for long :(

---

I knew this wouldn't stay a wiki free zone for long :(

Thanks for all the input. I will definitely update the app to include an initial % done to hopefully rid this problem. And will post the fact that it is beta now, but hopefully will not be soon.

I knew this wouldn't stay a wiki free zone for long :(

The good news is that you don't have to read it if you don't like it ...

---

Paul

Have you considered that as useful as the wiki is, that you are getting a bit carried away with it.

It is difficult for anyone to ask any questions on the boinc forum these days.
Instead of taking the time to help someone the lazy option is just to point to the wiki.

This behaviour is not unlike the RTFM mentality which put so many folks off linux.

---

The Wiki is there to help. I'm not a fan of it but it is very useful for new people to BOINC as BOINC doesn't come with an Instruction manual and can be daunting to some users.

I don't see it getting in the way of asking Questions

---

Join us in Chat (see the forum) Click the Sig


Join UBT

I also started my first WU. In my (rather slow) computer it takes 434 hours before october, 2nd. Even if I kept my computer running 24/7 until then, there are only 14*24=336 hours available. :-(
CPDN has long WU's too, but gives enough some time (350 days) to finish it.
This thread gives me some hope that the 434 hours will decrease when it's recomputed. I'll wait and see for the next days.

Roel

---

Have you considered that as useful as the wiki is, that you are getting a bit carried away with it.

Perhaps. But, what is the alternative? Not write anything at all?

It is difficult for anyone to ask any questions on the boinc forum these days. Instead of taking the time to help someone the lazy option is just to point to the wiki.

I don't think that is the case. In many cases, yes, for the common questions that get asked time and again, the easiest thing to do is to point to the Wiki. Or in the old days, my older site. A case in point that is common is Credit. This is not a simple topic. If it was it would be simple to explain. But, rather than repeat the same text over and over, we captured the essences and now they are in the Wiki for all to read.

When a new question comes up, or we find the material is not correct or incomplete we make changes.

This behaviour is not unlike the RTFM mentality which put so many folks off linux.

There are a couple people that spend a lot of time working the boards and helping people. I tend to lurk and comment only a little to save my time for writing. They are really pretty good with advice and usually they give a simple answer and then point to the Wiki as the complete answer. In many cases the person did not know that there was documentation.

Anyway, all I can say is that in general the comments are of the "keep it up" variety. And, again, you don't have to read it or refer anyone to it either.

---

Paul, Keep up the great work on the documentation! There's always room for questions in the message boards, but the Wiki is a great resource to have available to turn to.

---

I reckon quite a percentage of people get their answers from the Wiki and then post them as their own to make them look good anyway.

Keep up the good work....it looks great. I just wish I knew more so I could contribute to it.

---

David,

I was not thinking on quiting. :)

I was just asking for an alternative. I know we have so much in there and that it can be hard to find things, but, we do try to make it as easy as possible. And as correct as possible. But, to not have anything does not seem like a solution. Nor to limit the coverage to only BOINC, or ONLY production projects or only to "v" what ever "v" might be.

This is not the first time I have had someone tell me that having reference material is a bad thing, the product should work and be "intuitive" enough so that everything is clear. Since I don't know how that can be done ...

Anyway, I have to have some reason to get out of bed in the morning ...

Clark,

You *CAN* contribute. Read it and tell me where we were not clear. Look at the sections where we suggest a new person go first, tell us if that could be done better.

I only hang out in the NC forums, but, others send me e-mails on things they find in other boards. Just not enough time for me to be everywhere. And, you too can send me notes. [email protected]

---

Has this been fixed because I seem to be suffering the same effects.

Got 20 minutes in and no progress done.

Will leave it crunching a while to see what happens. My estimates seem fine...
2:39:43 hours...

Thats a AMD 64 3500+, 1Gb RAM, Nvidia 6200 TurboCache

---

Has this been fixed because I seem to be suffering the same effects.

Got 20 minutes in and no progress done.

Will leave it crunching a while to see what happens. My estimates seem fine...
2:39:43 hours...

Thats a AMD 64 3500+, 1Gb RAM, Nvidia 6200 TurboCache

The progress should go in large increments (8.3%) for this test protein we are running. It may take longer for this to show up in the progress.

---