In reply to Vitalii Koshura's message of 18 Apr 2026:

I have never seen this issue even when I have several project in the list that are not active anymore.

I noticed my WUProp task not reporting after a couple of minutes. I clicked Update and still no new task. I saw that GPUGrid was updating, and updating and updating. The log shows it was updating for 15 min and WUProp would not update while its waiting for another project. I detach G{UIGrid and WUProp and minecraft immediately update and download more work.


56533 minecrafthome 5/5/2026 4:13:21 AM Project requested delay of 7 seconds
56534 gpugrid 5/5/2026 4:14:12 AM Sending scheduler request: Requested by project.
56535 gpugrid 5/5/2026 4:14:12 AM Requesting new tasks for NVIDIA GPU
56536 minecrafthome 5/5/2026 4:16:03 AM Computation for task diaveins_1.00_469977_1 finished
56537 minecrafthome 5/5/2026 4:16:03 AM Starting task diaveins_1.00_469204_0
56538 minecrafthome 5/5/2026 4:16:05 AM Started upload of diaveins_1.00_469977_1_r279520193_0
56539 minecrafthome 5/5/2026 4:16:07 AM Finished upload of diaveins_1.00_469977_1_r279520193_0 (380 bytes)
56540 minecrafthome 5/5/2026 4:22:22 AM Computation for task diaveins_1.00_458357_1 finished
56541 minecrafthome 5/5/2026 4:22:22 AM Starting task diaveins_1.00_451892_1
56542 minecrafthome 5/5/2026 4:22:24 AM Started upload of diaveins_1.00_458357_1_r1907890813_0
56543 minecrafthome 5/5/2026 4:22:26 AM Finished upload of diaveins_1.00_458357_1_r1907890813_0 (420 bytes)
56544 minecrafthome 5/5/2026 4:25:22 AM Computation for task diaveins_1.00_469204_0 finished
56545 minecrafthome 5/5/2026 4:25:22 AM Starting task diaveins_1.00_469765_1
56546 minecrafthome 5/5/2026 4:25:24 AM Started upload of diaveins_1.00_469204_0_r1957168242_0
56547 minecrafthome 5/5/2026 4:25:26 AM Finished upload of diaveins_1.00_469204_0_r1957168242_0 (540 bytes)
56548 WUProp@Home 5/5/2026 4:26:24 AM Computation for task data_collect_v4_1777669201_164722_0 finished
56549 WUProp@Home 5/5/2026 4:26:26 AM Started upload of data_collect_v4_1777669201_164722_0_r837274_0
56550 WUProp@Home 5/5/2026 4:26:30 AM Finished upload of data_collect_v4_1777669201_164722_0_r837274_0 (30787 bytes)
56551 WUProp@Home 5/5/2026 4:28:27 AM update requested by user
56552 gpugrid 5/5/2026 4:29:32 AM Resetting project
56553 gpugrid 5/5/2026 4:29:32 AM Detaching from project
56554 WUProp@Home 5/5/2026 4:29:36 AM Sending scheduler request: Requested by user.
56555 WUProp@Home 5/5/2026 4:29:36 AM Reporting 1 completed tasks
56556 WUProp@Home 5/5/2026 4:29:36 AM Requesting new tasks for CPU
56557 WUProp@Home 5/5/2026 4:29:38 AM Scheduler request completed: got 1 new tasks
56558 WUProp@Home 5/5/2026 4:29:38 AM Project requested delay of 7 seconds
56559 WUProp@Home 5/5/2026 4:29:40 AM Started download of data_collect_v4_1777669201_167036
56560 WUProp@Home 5/5/2026 4:29:42 AM Finished download of data_collect_v4_1777669201_167036 (64 bytes)
56561 WUProp@Home 5/5/2026 4:29:42 AM Starting task data_collect_v4_1777669201_167036_0
56562 minecrafthome 5/5/2026 4:29:43 AM Sending scheduler request: To report completed tasks.
56563 minecrafthome 5/5/2026 4:29:43 AM Reporting 3 completed tasks
56564 minecrafthome 5/5/2026 4:29:43 AM Requesting new tasks for NVIDIA GPU
56565 minecrafthome 5/5/2026 4:29:45 AM Scheduler request completed: got 3 new tasks

I have done this exact same process several times. The client will only communicate with one project at once. For updates, for getting work for mass adding projects.

In reply to mmonnin's message of 5 May 2026:

The client will only communicate with one project at once. For updates, for getting work for mass adding projects.

And it has to do it that way for a very good reason- if it tries to contact multiple projects t the same time, how does it determine how much work to ask for, for each project?

You have your Resource share settings, cache settings, work preferences within projects.
In order not to miss deadlines, the Manager needs to request an amount of work that meets all of your preferences.
It will request it from one project, and once that has occurred it can then re-calculate whether the system can handle more or not without missing deadlines, If so, it will request work from the next project that has a work debt. Based on the result of that request, it'll do a request for the next project, etc, etc.



Have you made any changes to the network timeout settings?
Generally, my system will give up on a Scheduler request after 30-45 seconds or so if it doesn't get a response, certainly within a couple of minutes. The hard limit default is 5 minutes.

With Seti, I often had issues reporting a large number of results at once and i had to limit the number of Results reported at time in order for things not to time out.
Are your upload and download speeds with that project very fast or very slow. How large is the Scheduler file being transferred?

---

Grant
Darwin NT.

In reply to Grant (SSSF)'s message of 5 May 2026:

In reply to mmonnin's message of 5 May 2026:

The client will only communicate with one project at once. For updates, for getting work for mass adding projects.

And it has to do it that way for a very good reason- if it tries to contact multiple projects t the same time, how does it determine how much work to ask for, for each project?

Then separate getting work from getting preferences, reporting work, changing resource share or location, etc.
Instead of skipping over a project the client just does nothing.

It's not a new thing or anything to do with my setup. Devs think that unless they see it, then it can't happen. So there it is.

Not a Dev here but have dabbled with many projects over the years and every time I have had repeated backing off messages it has been one of these reasons.
1.Computer doesn't need work.
2.Project preferences don't allow the type of tasks the project has available.
3.Project has no tasks available.
4.My computer doesn't meet the specs or have the right operating system for the project.
5.Task failures mean the computer has reached its limit of tasks for the day.

None of these scenarios are the fault of the BOINC program.

In reply to mmonnin's message of 5 May 2026:

In reply to Grant (SSSF)'s message of 5 May 2026:

In reply to mmonnin's message of 5 May 2026:

The client will only communicate with one project at once. For updates, for getting work for mass adding projects.

And it has to do it that way for a very good reason- if it tries to contact multiple projects t the same time, how does it determine how much work to ask for, for each project?

Then separate getting work from getting preferences, reporting work, changing resource share or location, etc.
Instead of skipping over a project the client just does nothing.

Seriously?
Contacting the Scheduler is how all those things are done as it involves communication with the project.
Getting work can't be done until previously done work is reported. Why have two completely separate mechanisms for that???

And even having all of these things addressed through multiple separate mechanisms still won't stop this issue (or a similar one) from occurring, as whatever is causing things to stall and not time out is likely to be imbedded in all of these additional independent contact points with the project that you want to implement.
So while it won't block the other functions, it will block the function it is meant to be doing.

Have you made any changes to the network timeout settings?
Are your upload and download speeds with that project very fast or very slow. How large is the Scheduler file being transferred?

Answering the questions i asked could actually be of use in helping resolve the issue.

---

Grant
Darwin NT.

In reply to Grant (SSSF)'s message of 6 May 2026:

In reply to mmonnin's message of 5 May 2026:

In reply to Grant (SSSF)'s message of 5 May 2026:

In reply to mmonnin's message of 5 May 2026:

The client will only communicate with one project at once. For updates, for getting work for mass adding projects.

And it has to do it that way for a very good reason- if it tries to contact multiple projects t the same time, how does it determine how much work to ask for, for each project?

Then separate getting work from getting preferences, reporting work, changing resource share or location, etc.
Instead of skipping over a project the client just does nothing.

Seriously?
Contacting the Scheduler is how all those things are done as it involves communication with the project.
Getting work can't be done until previously done work is reported. Why have two completely separate mechanisms for that???

And even having all of these things addressed through multiple separate mechanisms still won't stop this issue (or a similar one) from occurring, as whatever is causing things to stall and not time out is likely to be imbedded in all of these additional independent contact points with the project that you want to implement.
So while it won't block the other functions, it will block the function it is meant to be doing.

Have you made any changes to the network timeout settings?
Are your upload and download speeds with that project very fast or very slow. How large is the Scheduler file being transferred?

Answering the questions i asked could actually be of use in helping resolve the issue.

Its NOT my issue. Its BOINC trying to contact a down project and getting stuck on it. I literally gave you a log and you wont believe it. Its not a one time instance. A prior time was when SiDock and Rakesearch went down. I could not get a new task until I removed SiDock from the client. The exact same situation.

In reply to mmonnin's message of 2 Jun 2026:

Its NOT my issue. Its BOINC trying to contact a down project and getting stuck on it. I literally gave you a log and you wont believe it. Its not a one time instance. A prior time was when SiDock and Rakesearch went down. I could not get a new task until I removed SiDock from the client. The exact same situation.

And still you don't answer the questions that may give some idea of what is happening.


An issue can't be fixed, if it can't be replicated.
There are many others here that are attached to all sorts of projects, yet they haven't come across this problem.,
So those that have the issue, need to help those that might be able to fix the issue, by supplying them with relevant information- such as what is different about their setup (which has a problem) compared to everyone else's (which don't have a problem).

---

Grant
Darwin NT.

No I have made no changes to network settings. What does that tell you? It doesn't matter as you don't even believe logs.

In reply to mmonnin's message of 6 Jun 2026:

It doesn't matter as you don't even believe logs.

If i didn't beleive the log, then i wouldn't have wasted my time and effort asking you questions about the issue, would i???

But as you obviously haven't paid the slightest bit of attention to what i have posted in relation to trying to understand this issue (which would be necessary in order to resolve it)- and here is the perfect example;

No I have made no changes to network settings.

That was one of three questions i asked you, and you continue to complain about things that aren't even true, then there is no point in me wasting further time in doing so.

---

Grant
Darwin NT.