I\'ve been running CPDN for a few days now and it\'s been running 9hours and 50 minutes but I still haven\'t gotten any credit at all ...

After how long do I get my first credit ?
I search the other forum and they said when 1 WU cycle is done or 10802 timestamps,
but how long is 1 timestamp and how do I know when a WU cycle is done ?

---

You get credits every time your computer uploads a trickle, which your\'s has recently done.
The actual allocation of credits is done by a program which runs every 12 hours.
Also, trickles are transferred from the receiving server to your account page by a program which runs every half hour, so it can sometimes take a short time to appear after BOINC has trickled.

There is some info on \"how often\" <a href=\"http://climateapps2.oucs.ox.ac.uk/cpdnboinc/quick_faq.php#l\"> here</a> in the FAQ.

Ok, I got it.

You will get credit for each trickle and there are 72 trickles in a model.

Timestep = 2.2 Seconds
Trickle = 6.6 Hours
Phase = 6.6 Days
Model = 19.8 Days

---

How ever, the above times assumes that you have a fast P4 computer.
And I have a half fast AMD64 3000+ , so the question is, how long is it on
such machine and is there a way to see if BOINC have returned any result for validation yet.

As of yet I can\'t find anything returned.
It has communicated with the server, saying

15/11/2005 14:56:14|climateprediction.net|Sending scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi
15/11/2005 14:56:14|climateprediction.net|Requesting 0 seconds of work, returning 0 results
15/11/2005 14:56:15|climateprediction.net|Scheduler request to http://climateapps2.oucs.ox.ac.uk/cpdnboinc_cgi/cgi succeeded


But it always says. \"returning 0 results\" never anything else.

How ever I can now see that I have gotten credit.
I have 94.52 in granted credit but it would still be nice to know the exact breaking point of each part you get credit.

Mostly because if you\'d like to run another project for a while
so you don\'t suspend the project when a Trickle is 99% done
or so and can estimate exactly when to switch over...

---

The model checkpoints every 144 timesteps. On the visualisation that\'s at 0030 every 3 model days (1, 4, 7, etc. of the month).

Each phase returns 24 trickles, which are a mechanism to let the project team know the model is still active and allow credits to be assigned before the model completes. Otherwise you would have to wait until the results are uploaded when all 3 phases of your current HadSM3 slab model have completed until you received any credits.

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

If you are using one the latest 5.* version of BOINC, you should get a message along the lines of: contacting the server to send a trickle.

Also, since you are running multiple projects, make sure that you set your preferences, (on your Account page), for: \"Leave applications in memory while preempted?\", to yes.
This way you won\'t have to re-run the data from the last check point when cpdn gets it\'s next allocating of computer time.

I\'m using BOINC 4.45 since the 5.* versions doesn\'t
support the optimized setiathome_4.11_windows_intelx86.exe client.
At least I haven\'t being able to make it work, I just get client errors. :o/


But ok, I got my fist credits according to my account page after 29115 sec,
which should indicate that a trickle is that long for me,
so 2 should be 58230 sec (16h18m) , 3 = 87345 sec (24h27m) etc. etc ...


So I could really calculate it from there, how ever I wonder if it reports
the Trickle immediately when I click on the \"Update\" button
or if it has certain time intervals that it reports the Tricklets ...

---

---

Also, since you are running multiple projects, make sure that you set your preferences, (on your Account page), for: \"Leave applications in memory while preempted?\", to yes.
This way you won\'t have to re-run the data from the last check point when cpdn gets it\'s next allocating of computer time.

---

Hm, for the moment I\'m running 7 projects on one computer and
I can\'t leave them all in memory, it would make my computer lag eventually.

Also I switch between the projects manually by suspending all but
the one I want to run at the moment as I getting an error that says:

\"Using earliest-deadline-first scheduling because computer is overcommitted.\"

If I leave it auto it crunches all Wus for just one project, before starting
on any WU for any other project.

This is due to that the every Project has different size WUs and the one that has the
smallest WUs and therefor has the shortest deadline times and they are therefor crunched first.

---

If you are permanently connected to the internet, and you DON\'T have \"Disable BOINC network access\" selected, (under File in the menu), then trickles will upload when they are created. There is no need to do anything.
Otherwise, they will upload the next time you connect, and / or enable \'net access. Again, no need to do anything else.

An Update will immediately try to upload anything waitng to be uploaded, and download the latest preference changes, (if any), as well as the latest value of your credits.

edit
data kept in memory will be swapped to your hard disk when and if necessary. It will be swapped back when needed.

\"Earliest deadline first\" is NOT an error. It is just normal behavour for BOINC!
Perhaps yopu should read <a href=\"http://boinc-doc.net/boinc-wiki/index.php?title=Work_Scheduler\"> this.</a>
It seems as though you are being overly ambitious in running so many projects, if you have to manually switch between them.

Well, that is just your opinion.

I personally think that the deadlines are extremely to narrow.

Just think of the people who only uses their computer a few times per week,
they always miss all deadlines and haven\'t got a chance of getting any credit.

In that way I think the old SETI@Home classic was best.
No deadlines at all !!!

And besides infact my computer isn\'t over committed, I have a very fast computer
a AMD64 3000+ with 512mb DDR400 SSE2 memory and it will make all WUs in the time limit.

But for some reason BOINC doesn\'t think it will, why I don\'t know ...

But ok, I changed the \"Leave applications in memory while preempted? \" to yes now... perhaps it will help some...

---

In that way I think the old SETI@Home classic was best.
No deadlines at all !!!

Sorry but seti classic had deadlines too. After about a month a returned result was simply put in the bit bucket.

---

BOINC WIKI

BOINCing since 2002/12/8

In that way I think the old SETI@Home classic was best.
No deadlines at all !!!

Sorry but seti classic had deadlines too. After about a month a returned result was simply put in the bit bucket.

Perhaps I don\'t know, but that was still a month, now we have about 2 weeks at best...

---

The great thing about BOINC is that users have so much choice. People who run their computers for only a few hours a week, or who have slow machines, obviously will not be able to get through much work and so will probably only sign up to one project at a time. Others are in a very different position. At the extreme, Carl from this project has asked for volunteers with 3.2 GHz P4 processors or better to run the spinup experiments full time for up to four months. That is not suitable for many people, but there are other things that are.

It is true that BOINC is still having trouble estimating how long a WU will take on each machine, but thats gets better with time and I don\'t think that deadlines can be accused of being tight on this project.

---