4/21/2022 8:24:58 PM | WUProp@Home | Sending scheduler request: Requested by user.
4/21/2022 8:24:58 PM | WUProp@Home | Reporting 1 completed tasks
4/21/2022 8:24:58 PM | WUProp@Home | Requesting new tasks for NVIDIA GPU
4/21/2022 8:24:59 PM | WUProp@Home | Scheduler request completed: got 0 new tasks
4/21/2022 8:24:59 PM | WUProp@Home | Not accepting requests from this host

This is my main system. The errror started happening within the last 24 hours.

This week, to see how much power I can conserve, I stopped running computational tasks and am only running NCI tasks. I do this every summer when the weather gets warm. I've never had a problem before.

Any ideas about how to correct this?

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

The host is banned. WUProp doesn't allow clients running only NCI apps. Otherwise we could setup 100s of clients to get 100k hours on the NCI apps. You'll have to run something w/in the 1 hour window of a WUProp task. Could even setup a schedule if needed to run 1 task for 5 minutes.

Does the host show up in the "Hosts running NCI apps W/O sensors" link on your My Account page?

The host is banned. WUProp doesn't allow clients running only NCI apps. Otherwise we could setup 100s of clients to get 100k hours on the NCI apps. You'll have to run something w/in the 1 hour window of a WUProp task. Could even setup a schedule if needed to run 1 task for 5 minutes.

I know about that. This is why I mentioned that I've been doing this for years.

Does the host show up in the "Hosts running NCI apps W/O sensors" link on your My Account page?

No it does not.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

Michael are you running the Radioactive app? It is not nci but takes almost no cpu time since their is no 'detector' attached, and bonus you get wuprop credits for running it but of course zero Radioactive Project credits.

Michael are you running the Radioactive app? It is not nci but takes almost no cpu time since their is no 'detector' attached, and bonus you get wuprop credits for running it but of course zero Radioactive Project credits.

I am not running it.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

The host is banned. WUProp doesn't allow clients running only NCI apps. Otherwise we could setup 100s of clients to get 100k hours on the NCI apps. You'll have to run something w/in the 1 hour window of a WUProp task. Could even setup a schedule if needed to run 1 task for 5 minutes.

I know about that. This is why I mentioned that I've been doing this for years.

Does the host show up in the "Hosts running NCI apps W/O sensors" link on your My Account page?

No it does not.

Yep, those tasks w/o a compute task used to be invalid. Now they seem banned.

It is not nci but takes almost no cpu time

NCI = non CPU intensive. Radioactive is NCI even if it doesn't have the label. Example: iThena

The host is banned. WUProp doesn't allow clients running only NCI apps. Otherwise we could setup 100s of clients to get 100k hours on the NCI apps. You'll have to run something w/in the 1 hour window of a WUProp task. Could even setup a schedule if needed to run 1 task for 5 minutes.

I know about that. This is why I mentioned that I've been doing this for years.

Does the host show up in the "Hosts running NCI apps W/O sensors" link on your My Account page?

No it does not.

Yep, those tasks w/o a compute task used to be invalid. Now they seem banned.

It is not nci but takes almost no cpu time

NCI = non CPU intensive. Radioactive is NCI even if it doesn't have the label. Example: iThena

I am still getting hours in wuprop for them, this is since yesterday:

Radioactive@Home Radioactivity Monitor 27,786.78 43.92

Yep, those tasks w/o a compute task used to be invalid. Now they seem banned.

If that's true, anytime you are running just one project and it runs out of work, your computers will get banned.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

After about two days, it started working again all by itself.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

That error usually means that no project was detected as running so the work unit becomes invalid.

When you get too many errors the host is stopped from getting more work till a time out is reached and each successful work unit allows more work to be sent out.

NCI work should still be valid, you should be able to run something like iThena without other projects as long as you don't run more BOINC instances than the number of cores that your computer has and try to get WUProp hours for them.

This was something that Sebastien set up a few years ago.

That is how I understand it.

Conan

---

Since the restart of the projects, 25% of returned results were invalid.
To prevent too many invalid results, I introduced a temporary ban for hosts which returns too many invalid results.

An host which returns 12 or more invalids results in 24 hours will be banned for 24h. In case of recidive, bannish time will be longer.

You can know the ban time limit visiting host details. For example: https://wuprop.boinc-af.org/show_host_detail.php?hostid=154599

For information, a result is invalid if only wuprop project runs on this host. Running only NCI applications is allowed.

Since the restart of the projects, 25% of returned results were invalid.
To prevent too many invalid results, I introduced a temporary ban for hosts which returns too many invalid results.

An host which returns 12 or more invalids results in 24 hours will be banned for 24h. In case of recidive, bannish time will be longer.

You can know the ban time limit visiting host details. For example: https://wuprop.boinc-af.org/show_host_detail.php?hostid=154599

For information, a result is invalid if only wuprop project runs on this host. Running only NCI applications is allowed.

The problem (from the perspective of gathering data about projects with a low availability of tasks) is that a computer which is set to run only projects with intermittent work will find itself locked out of WUProp. Then when those intermittent tasks are run, WUProp isn't running.

---

Want to find one of the largest known primes? Try PrimeGrid. Or help cure disease at WCG.

Adding to this problem is the fact that WUs are now only 1 hour in duration. Previously, they were 6 hours.

Sebastien,

I tend to call myself a long time supporter, and we had contact several times the past 12+ years, but to be honest, I think it's an unfortunate if not stupid decision. I - as well as many others for sure - have loads of projects involved that don't have a constant supply of enough work, so the computer runs dry more often than not. To ban the computer because of this is just stupid.

Very unfortunate, but I will need to disconnect my hosts from WUprop if you decide to keep it that way. I don't have the will to support nonsense.

BR
aend

I'm testing a new version of application which should prevents return of results without data.

Before the workunit will end, application will check if result contains data. If there is no data, workunit end will be postponed until result contains data.

This computer has been banned until 21 May 2022, 16:14:25 UTC

My favorite project sends WU's again, is there anything I can do to make WUprop count the computing times again? I'm sorry for supporting such an unreliable project, and sometimes I don't get WU's...

My Mac M2, because of heat issues, I only run intermittently, for challenges and the like. I use "no new work" to do this - but I've always just left WuProp running. Now I find it's banned until "---" (no date in field) - is there a "reset host" button somewhere for manual unbanning, when we've "fixed the problem" (or at least now know there IS a problem)?

I've reset, removed and reattached. Maybe change the name and create a new host ID?

Nope. Changed host name, still logged in as same host ID, name change didn't propagate. :shrug:

My Mac M2, because of heat issues, I only run intermittently, for challenges and the like. I use "no new work" to do this - but I've always just left WuProp running. Now I find it's banned until "---" (no date in field) - is there a "reset host" button somewhere for manual unbanning, when we've "fixed the problem" (or at least now know there IS a problem)?

I've reset, removed and reattached. Maybe change the name and create a new host ID?

Set wuprop to no new tasks as well and then when you start your work back up again restart wuprop as well. Also once you start running tasks again and you have wuprop set to send you new tasks they will automatically start coming again once your ban is over.

The computer isn't banned anymore. To avoid issues like that in the future, follow mikey's advice and set WUProp to no new tasks together with the other projects.

---