From BOINC Manager:

12.07.2025 14:49:07 | Asteroids@home | Requesting new tasks for CPU
12.07.2025 14:49:10 | Asteroids@home | Scheduler request completed: got 0 new tasks
12.07.2025 14:49:10 | Asteroids@home | Project has no tasks available
12.07.2025 14:49:10 | Asteroids@home | Project requested delay of 121 seconds

Server status: 12.07.2025 14:54:00
Tasks ready to send 1131079

Seconded...

"C'mon man..." - Frank Drake

Tasks ready to send 1135260
Project has no tasks available
Possible they are limiting the number of computers?

I don't believe there's a deliberate limit. The work buffer may be just too busy due to the high volume of short tasks.
We've told Radim about the issue. Hopefully, he’ll be able to fix it soon...

I'm getting this message now for my GPUs, I don't know if CPUs aren't being handed out too, but I see a third of a million tasks available on the server status page. So where are they? Are they being pre allocated to GPU CPU or mobile and then one queue running out? This needs to be sorted, my GPUs are sat idle.... (well actually they run Einstein instead, but you're losing out here).

I'm getting this message now for my GPUs, I don't know if CPUs aren't being handed out too, but I see a third of a million tasks available on the server status page. So where are they? Are they being pre allocated to GPU CPU or mobile and then one queue running out? This needs to be sorted, my GPUs are sat idle.... (well actually they run Einstein instead, but you're losing out here).

imagine you have a 100,000 L water tank. but you cant take water from the tank directly. instead you need to get water from a small 10L pail that fills up from the tank. now imagine that 1000 people are trying to get water from that small pail all at the same time. most people will only get a small amount or nothing depending on when they go to fetch from the pail. that seems to be the basis of the problem. the pail is too small.

tasks are not pre-allocated to any particular device. it's first come, first serve.

---

This does not seem to be such an issue at other projects so why is it here?
Is it a config or resource issue?
Still getting "no tasks available".

---

Paul.

I'm getting this message now for my GPUs, I don't know if CPUs aren't being handed out too, but I see a third of a million tasks available on the server status page. So where are they? Are they being pre allocated to GPU CPU or mobile and then one queue running out? This needs to be sorted, my GPUs are sat idle.... (well actually they run Einstein instead, but you're losing out here).

The project is all out whack right now.
I was just able to upload and report 14 tasks.
It could be in part that the various servers are working to get the backlog taken care of and that all of us are pounding the system at the same time looking for work.

Give it 8-12 hours to get any last bugs out and for the system to stabilize back to normal.
They have been down for a bunch of days at least from what I was seeing here.

Tasks ready to send 1135260
Project has no tasks available
Possible they are limiting the number of computers?

It is possible that they are really limiting the number of counters or some accounts.
I got this message today when trying to upload new tasks.

29.07.2025 5:30:32 | Asteroids@home | Not requesting tasks: don't need (CPU: ; NVIDIA GPU: blocked by project preferences)

Tasks ready to send 801017
Tasks in progress 108204
Workunits waiting for validation 27009

I think they are trying to play catch up.
And those of us that have sent work back recently are not getting anything new.
I also have no new tasks, but I dumped 14 or so on them after they got running again.

But I am focusing on getting my LHC RAC back after something went nuts in my virtualization setup, so if I don't get any work here its not a problem.

> @monty Not requesting tasks: don't need
If your client already has enough tasks from other projects (for example, 1-2 days based on your preferences). It won’t request additional ones.

Between 27.7-28.7 the server ran out of disk space, which has been resolved

However, the original issue persists. The server does not send (or rarely sends) new work when the client requests it, even though work is available. The administrator is aware of the problem, but its cause is still unclear. If there is a limit or misconfiguration, it is unintentional.

I have new work.
The validator has now caught up, that may have been a factor in lack of work sent.

---

Paul.

new work ended quickly ;o)

This does not seem to be such an issue at other projects so why is it here?
Is it a config or resource issue?
Still getting "no tasks available".

The fact that it is this project in particular makes it that much more unsettling...for some reason the intro to Lifeforce keeps popping into my head. Cheers.

Its break time at the universities. Maybe we will see an increase in work in september.

new work ended quickly ;o)

My PCs are resorting to Einstein. I just allocate a weight to each project depending how interesting it is, and let Boinc find work anywhere. I'm on every project, but some are classed as a million times more important :-)

The fun part is adjusting the app_config files so any combination of work maximises the GPUs and CPU.

here its bone dry. Rosetta is bone dry. GPU went dry on stuff that a 1080 can work on and now the big expensive gpus have work.

Wait a month and see what happens. At least LHC and Einstein keep steady. And wcg is chugging away. I am on all the projects there just to keep credit up

here its bone dry. Rosetta is bone dry. GPU went dry on stuff that a 1080 can work on and now the big expensive gpus have work.

Wait a month and see what happens. At least LHC and Einstein keep steady. And wcg is chugging away. I am on all the projects there just to keep credit up

try GPUGRID on the 1080 again. they have a bunch of ATM tasks which should run on that card. it says 0 RTS, but they are on a revolving door (new ones are created when old ones returned) so you just have to keep asking for new work to get them.

---