Website: Tasks ready to send = 806037

Boinc Client: Fr 08 Mär 2024 23:03:04 CET | Asteroids@home | Project has no tasks available

What is the explanation for this little difference?

In best case I get 6 WUs after dozens of requests.

Nobody knows. I get the same thing. Half a dozen requests for work and get nothing and finally one succeeds.

Doesn't help the project was chosen for a Formula Boinc contest the other day either. Scheduler is getting hammered more than usual.

---

A proud member of the OFA (Old Farts Association)

"Tasks ready to send" is the number of tasks that have been generated (so they are in the database) but not sent yet. However, the scheduler works with its own job cache, not with the full task table in the database. This cache can be much smaller, and if it happens to be empty when a client requests work, the client will get the "Project has no tasks available" message. Also, not all tasks from this job cache may be suitable for every client, as there are mechanisms such as homogeneous redundancy or accelerated retries that may impose some constraints, or a project may choose not to send more than one task of each workunit to the same client or user.

Usually, this is not much of a problem here, because the backoff time between two requests from the same client is only a few seconds, so the client can retry and eventually get tasks quite quickly. But if more clients compete for the same job cache, it will be empty more often, and the short backoff time then leads to even more unsuccessful retries. This likely goes a long way explaining the problems in the past few days (except for the expired SSL certificate yesterday). So it might help a lot if the scheduler's job cache was refilled more frequently.

Nobody knows. I get the same thing. Half a dozen requests for work and get nothing and finally one succeeds.

Doesn't help the project was chosen for a Formula Boinc contest the other day either. Scheduler is getting hammered more than usual.

The Formula BOINC Challenge has selected Asteroids as a project in the past without issues...so, if this is putting more strain on the servers, then the servers should easily be able to cope with some extra demand, as they have done so in the past.

sorry for the trouble, the server should easily handle +200 (/4500) more users during the challenge. Perhaps the webserver ran out of free connections or more likely someone's deliberately hitting it too hard (by downloading an insane amount of work, crawling the website etc.). Unfortunately, only the university has access to the firewall & logs and it may take some time for them to investigate what exactly has happened especially during a weekend. They're aware of the issue, but I don't know more yet.

I have some news from admins, there was a DOS attack from China's network on the website 1000 to 8000 connections at a time. They took some preventive measures. It's hard to say if there's any connection with the Challenge though, it could be just a coincidence.

Thank you for the info, this can certainly explain the webserver issues. However, it cannot explain why the scheduler cache is empty quite often despite tasks being ready to send (and this also happened before the challenge, albeit less frequently). Maybe the feeder is just slow because of the large number of tasks in the database, but maybe something can be tweaked to make it faster even with a large database.

I have some news from admins, there was a DOS attack from China's network on the website 1000 to 8000 connections at a time. They took some preventive measures. It's hard to say if there's any connection with the Challenge though, it could be just a coincidence.

Hiya

Thanks for the information.

I agree it looks like a coincidence, as the number of Formula BOINC members taking part in the Challenge is relatively small (although I cannot say how many hosts each member is using !!).

regards
Tim
Admin, Formula BOINC Sprints

Maybe the feeder is just slow because of the large number of tasks in the database, but maybe something can be tweaked to make it faster even with a large database.

Hi

I agree with the above - I have tasks in my "Valid tasks" list going back to 19th November 2023.

These have been awarded credits and by now should have been purged from the database?

And both have met the required quorum as well !

Maybe someone can advise the admin to take a look at this issue?

Is this still a problem? My old AMD cards (Tahiti, 280X) will not get work, they used to. Or has the new program become incompatible with them? I'm only getting CPU work.

Read this thread. https://asteroidsathome.net/boinc/forum_thread.php?id=967#7970
Seems the 280X is only OpenCL 1.1 capable.
Check your Boinc startup log for the gpu detection output. If it only reports OpenCL 1.1 you'll have to use it elsewhere.

---

A proud member of the OFA (Old Farts Association)

Doesn't seem to be true, they're getting them now. An old dual core CPU was starved too, I think it's a case of waiting for the right type of tasks to appear. They appear faster than they can be processed, it's just an initial delay.