Hey,
I joined malariacontrol a few days ago. At first, my work units would get sent back immediately after completion, but now they don't. When a work unit is finished, it will say "uploading" but then it just becomes "ready to report". It won't report until I manually click "update" in the BOINC window. I had it set to 0 for how often I connect to the network, and it still isnt working. I tried setting to a really small but positive number, like 0.02 (30 minutes or so) but the work units still won't report. Though, they say uploading. But, they don't appear in my pending credit list or anything. Maybe I don't totally understand how this works, but this isn't normal, right? Why arent the work units being Reported? I could have sworn it was working properly when I first started. Thanks for the help!

Hey,
I joined malariacontrol a few days ago. At first, my work units would get sent back immediately after completion, but now they don't. When a work unit is finished, it will say "uploading" but then it just becomes "ready to report". It won't report until I manually click "update" in the BOINC window. I had it set to 0 for how often I connect to the network, and it still isnt working. I tried setting to a really small but positive number, like 0.02 (30 minutes or so) but the work units still won't report. Though, they say uploading. But, they don't appear in my pending credit list or anything. Maybe I don't totally understand how this works, but this isn't normal, right? Why arent the work units being Reported? I could have sworn it was working properly when I first started. Thanks for the help!

Thanks for your help.

[quote]
Completed work is reported at the first of:

1) 24 hours after completion
2) 24 hours before report is due.
3) Connect every X before report is due.
4) On a trickle up message (CPDN only so far).
5) On a request for more work.
6) On a manual update.

Hope this helps

Michael

[quote]
Completed work is reported at the first of:

1) 24 hours after completion
2) 24 hours before report is due.
3) Connect every X before report is due.
4) On a trickle up message (CPDN only so far).
5) On a request for more work.
6) On a manual update.

Hope this helps

Michael

Is that how all the projects work, or does each project have its own characteristics? I've noticed Seti connects to the network, reports finished WUs, and gets new ones, fairly regularly, but some of the other other ones I run, namely Malaria and Einstein, do not.

[/quote]
The core client handles the reporting, so it is the same for all projects. The differences in how you are seeing the projects work will be due to the way your project preferences, project debts and the connection triggers are interacting.

Michael[/quote]


I see. So does debt accumulate for projects that you arent attached to yet? What I mean is, I signed up for Malaria, and it was my only project for almost a week, and I completed a good deal of work. Then, just a few days ago, I joined three more projects. Do I have a lot of debt for these to make up for all the time I spent solely on Malaria? It sure seems these other three are getting all of the work, as Malaria has not asked for more work units in a while, even though it has the highest resource share.

I see. So does debt accumulate for projects that you arent attached to yet? What I mean is, I signed up for Malaria, and it was my only project for almost a week, and I completed a good deal of work. Then, just a few days ago, I joined three more projects. Do I have a lot of debt for these to make up for all the time I spent solely on Malaria? It sure seems these other three are getting all of the work, as Malaria has not asked for more work units in a while, even though it has the highest resource share.

[I have noticed a post about completed work units not uplaoding. I have had this several times of late. I appear to have this problem though on all my projects! (6 of them). Also when I check the BOINC stats All Projects, none of my completed work units appear to be recognised, I am going backwards instead of forwards. I am probably in the wrong message board but I am new to this so please excuse my ignorance. Regards, and thanks for any help you can give me, Malcolm, ().]

[I have noticed a post about completed work units not uplaoding. I have had this several times of late. I appear to have this problem though on all my projects! (6 of them). Also when I check the BOINC stats All Projects, none of my completed work units appear to be recognised, I am going backwards instead of forwards. I am probably in the wrong message board but I am new to this so please excuse my ignorance. Regards, and thanks for any help you can give me, Malcolm, ().]

Hi, I have a problem... I don't get workunits for malariacontrol anymore. Did I miss something in the news? The last workunit was uploaded around May 6th, 2012, and from that date on I haven't received anything.
Can someone give me a clue what I might have done wrong?
Kind regards,
Roland
;-)

@Roland : Have you limited the workunit type to a specific Malaria application?
It can be changed in the project settings, the list of settings that start with "Run ..."
______________

As of the reporting :

7) If a project-enforced contact is due (next_rpc_delay) - Afaik. only WCG and RNA so far

Hi Ananas (?),

in reply to you message: I have these settings...
Run malariacontrol simulation application Yes
Run openMalaria test application Yes
Run map predictor application Yes
Run optimizer application Yes
Run openMalaria A application Yes
Run openMalaria B application Yes

In the past malaricontrol was halted due to a task of Climate prediction with a higher priority, but after a while malaricontrol took control again. But now I don't even have tasks for MC downloaded!

So, any help is welcome!

Kind regards,
Roland
;-)
____________

Is there no message at all, why it didn't get work? Does it even request work?

If it is something like :

Sending scheduler request: To fetch work. Requesting 3184 seconds of work, reporting 0 completed tasks

it did request work, but if it says "Requesting 0 seconds", it is probably either set to "no new work" or the local cache is so full from other projects that the computer does not need new work.


The project work supply can be checked here (Results ready to send 2,054 - so we did not run out of work projectwise)

Hi Ananas,

thanks for replying!

Here are a couple of lines from the "gebeurtenissen logboek" (the log):
...
26-5-2012 14:50:22 | malariacontrol.net | URL https://malariacontrol.net/; Computer ID 444528; resource share 100
...
26-5-2012 16:30:34 | malariacontrol.net | update requested by user
26-5-2012 16:30:36 | malariacontrol.net | Sending scheduler request: Requested by user.
26-5-2012 16:30:36 | malariacontrol.net | Requesting new tasks for NVIDIA
26-5-2012 16:30:38 | malariacontrol.net | Scheduler request completed: got 0 new tasks
26-5-2012 16:30:38 | malariacontrol.net | No work sent
26-5-2012 16:30:38 | malariacontrol.net | No work is available for malariacontrol.net
26-5-2012 16:30:38 | malariacontrol.net | No work is available for openMalaria test version
26-5-2012 16:30:38 | malariacontrol.net | No work is available for Prediction of Malaria Prevalence
26-5-2012 16:30:38 | malariacontrol.net | No work is available for Estimation of parameters of infection dynamics (variable duration, max 4h)
...

As I understand the client requests workunits, but doesn't get them. Now the question is: aren't there any workunits on the server, or won't the server send workunits because of settings on the client...

Kind regards,
Roland
;-)

I guess, this is the hint :

...
26-5-2012 16:30:36 | malariacontrol.net | Requesting new tasks for NVIDIA
...

Hi Ananas,

mmm, you could be right, but:

--- settings ---
Gedeelde bronnen
Bepaalt de hoeveelheid gedeelde bronnen van computers verbonden met dit project. Voorbeeld: Als je computer deelneemt aan 2 BOINC projecten met gedeelde bronnen van 100 en 200, zal het eerste project 1/3 en het tweede project 2/3 van de beschikbare bronnen krijgen. 100

>>>> Gebruik CPU Instelbaar vanaf 6.10+ ja <<<<

Test applicaties uitvoeren?
Dit helpt ons bij het ontwikkelen van applicaties, maar kan fouten veroorzaken tijdens het uitvoeren nee
Wil je dat malariacontrol.net en/of je team e-mail naar je kunnen verzenden?
Er worden door e-mails naar je verzonden; zorg ervoor dat een eventueel spamfilter ze niet blokkeert. ja
Wil je dat malariacontrol.net jouw computers laat zien op de website? nee
Standaard computer locatie home
Run malariacontrol simulation application Yes
Run openMalaria test application Yes
Run map predictor application Yes
Run optimizer application Yes
Run openMalaria A application Yes
Run openMalaria B application Yes
---

So, your suggestion is already active!

kind regards,
Roland
;-)

Hmmm, then maybe someone who knows more about later BOINC clients needs to jump in, I'm a bit out of ideas now, sorry :-/

Hi Ananas,
well, anyway, thank you very much for your help so far.
Kind regards,
Roland
;-)

Guess what... I asked ClimatePrediction to temporarily back off, and bingo: I received several "marlatia" files to process...

ClimatePrediction was/is busy with a high priority task, and I think this is the reason why I didn't receive files for MalariaControl.

I think & hope: problem solved.

Bye,
Roland
;-)

Ah, so your client version has two cache handlers.

That means that the local CPU cache has been stuffed but the GPU cache had still capacity and that's why it asked only for nVidia work. When you suspended the panic mode result, more CPU cache capacity has (temporarily) been freed.

I often suspend work in order to enforce new requests but I didn't expect that it had a separate effect on CPU and GPU requests.

Ah, so your client version has two cache handlers.

That means that the local CPU cache has been stuffed but the GPU cache had still capacity and that's why it asked only for nVidia work. When you suspended the panic mode result, more CPU cache capacity has (temporarily) been freed.

I often suspend work in order to enforce new requests but I didn't expect that it had a separate effect on CPU and GPU requests.

Hi Ananas,

MalariaControl works on CPU at my machine, as well as ClimatePrediction. The only project which runs on de GPU is Seti@home.
Do I understand it well from your response that MalariaControl might run on GPU as well?

kind regards,
Roland
;-)

Of course it might do, some day, but currently no.

But if the GPU cache is empty and the BOINC client has been told to fetch work, it will accept GPU work if it's allowed to by your settings. If both (CPU and GPU) caches are empty, it will request both types of work, even if the project is unable to supply it.

There are a bunch of GPU projects besides SETI, like MilkyWay, Collatz, Primegrid, afaik. even POEM has a GPU client now.


p.s.: you can see if a project supports GPU work, it's listed on the "Applications" link of each project (here that would be : https://malariacontrol.net/apps.php)