BOINCGames.com Super Bowl Sprint - Einstein@Home

[Skillz]

The first Sprint of the year has started. You need to register on boincgames.com with your CPID if you want to participate. It can take a little while for your CPID to get validated. I recommend changing your name at Primegrid, Universe@home, SRBase, AmicableNumbers or Universe@home as those projects update their stats export file multiple times a day with Universe@home having no work available for the foreseeable future.

The Sprint project is Einstein@Home.

If you are already registered for BG then all you need to do is switch GPUs to einstein@home and start crunching. Otherwise, register for BG and go ahead and move to Einstein@Home so when your CPID gets validated you can quickly start earning points.

 

[crashtech]

Oops! I got the date mixed up, but I'm on it!

 

[Skillz]

We're currently in 3rd, but we're clawing our way to the top. Come on and join in.

 

[StefanR5R]

These sprints are good for learning about project quirks.

Sun 11 Feb 2024 04:19:47 AM CET | Einstein@Home | Sending scheduler request: To fetch work.
Sun 11 Feb 2024 04:19:47 AM CET | Einstein@Home | Reporting 1 completed tasks
Sun 11 Feb 2024 04:19:47 AM CET | Einstein@Home | Requesting new tasks for NVIDIA GPU
Sun 11 Feb 2024 04:19:49 AM CET | Einstein@Home | Scheduler request completed: got 1 new tasks
Sun 11 Feb 2024 04:19:49 AM CET | Einstein@Home | Project requested delay of 60 seconds

These "report one result, get one task" conversations between client and server were the standard. That is, clients received rarely more than on task. On a fast host, therefore make sure that you have enough clients sharing the hardware, so that each client does not finish more than one task per minute.

And then,

Sun 11 Feb 2024 04:20:49 AM CET | Einstein@Home | Sending scheduler request: To fetch work.
Sun 11 Feb 2024 04:20:49 AM CET | Einstein@Home | Requesting new tasks for NVIDIA GPU
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | Scheduler request completed: got 0 new tasks
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | No work sent
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | No work is available for Binary Radio Pulsar Search (MeerKAT)
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | (reached daily quota of 1024 tasks)
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | Project has no jobs available
Sun 11 Feb 2024 04:20:50 AM CET | Einstein@Home | Project requested delay of 74836 seconds

What, 74836 seconds request delay? That's 20h 47m 16s.

(The daily quota of 1024 tasks is not the number of task in progress, it is the maximum number of tasks which the server assigns to a host during a 24h hour period. I do not know which hosts are put under this limit; at least new hosts have it, but not only them.)

In other words, at 4AM the server instructed the host to shut down all scheduler communications, and thus it has been involuntary bunkering results since then, besides not fetching new work. Well, I did not plan to bunker, but Einstein@Home apparently wants me to. ;-)

As there is a nonzero need for attention to this project, it is good that this sprint started on a weekend = not during the workweek.

 

[Skillz]

E@H has the ability to see what your client requests also.

If you go to your hosts detail page and look for this:



You can click the date and it'll show information such as this.

 

[StefanR5R]

Thanks. I only vaguely remembered this feature and had only be looking at local logs with debug flags enabled.

Gonna post the weekly stats, then reconfigure the clients to append the keywords pretty please to work requests.

 

[Skillz]

We are 12.7M points behind GPUUG, just need a little more to catch them before the end.

 

[Skillz]

Current members participating:



Current position:


Currently 12,398,463 points behind first place.

 

[Pokey]

I have a rig that is being cantanerous. I'm not able to go full out, for having to trouble shoot every five minutes.....................

 

[Skillz]

What's the issue?

 

[Pokey]

I had a panic attack.................
Took a while to figure out. I think the M.2 drive got corrupted. Switched to a SATA drive and things seem to have settled down. I hate computers.........................

 

[StefanR5R]

One of my computers froze at 3AM. It's too early in the morning (6AM) to figure out what happened; I just rebooted it.

 

[Skillz]

Unless GPUUG has a huge bunker (I don't think they do) or they add a ton of hardware in the next few hours, I believe we got this one. Great job team.

 

[StefanR5R]

The final stats update (visited at 12:15 UTC) shows 0 for "last 1 hour update" for all teams and all individuals. Also I believe, but am not sure, that "total credits for the sprint" are the same as after the 11:00 UTC update. Grafana shows that the project server is being up and has had good response time around 12:00 UTC. A glitch in the Boinc Games data acquisition? (Would post at the BG forum, but don't have my login credentials with me.)

 

[Skillz]

Yes, that's correct. Something is preventing the last update to go through. It's been added to the to-do list.

 

[JC]

Nertz, I missed it!!!

 

[Ken g6]

The next sprint is up and it's Milkyway. They don't do GPU work anymore; just multi-threaded CPU work.

Edit: If you haven't run it recently, you should also reset the project.

 

[Ken g6]

Milkyway is having problems with their Orbit Fitting WUs. This just started within the past day or so. Their other WU type seems fine.