Rally Apr. 17th-19th

[Mini-Geek]

Originally Posted by IronBits

When my KoW meter worked (just noticed it's dead now)...

Too much spinning for it to handle huh?

[IronBits]

Don't know, it used to work, who knows, it may have been dropped on the floor with all the spring cleaning.
No matter, they don't cost that much, I'll get another.

[mdettweiler]

Rats! I woke up this morning and found that all four cores of my quad are frozen at 99%! (Stupid bug! ) I hope I have time to get them restarted before I have to head out shortly...

Edit: Hmm, another look at the progress report tells me that they've been frozen since BEFORE the rally started (they froze at ~7AM 4/17 server time). And I didn't notice it?!?

[gd_barnes]

Froze at 99% of what?

[mdettweiler]

Originally Posted by gd_barnes

Froze at 99% of what?

99% on the progress bar. You know--the classic Windows 99% client bug. I'm sure you've encountered it before. (On Linux it often manifests itself as a 100% complete bug instead, which on the terminal console usually just shows the last result as the last line, and nothing after that with the client completely idle.)

Anyway, they're all back online now.

[Mini-Geek]

Originally Posted by mdettweiler

99% on the progress bar. You know--the classic Windows 99% client bug. I'm sure you've encountered it before. (On Linux it often manifests itself as a 100% complete bug instead, which on the terminal console usually just shows the last result as the last line, and nothing after that with the client completely idle.)

Anyway, they're all back online now.

Oh, you mean the status of the prime tests all get to 99% and then stop? I wasn't sure what you meant either, but I think I know now.

[mdettweiler]

Originally Posted by Mini-Geek

Oh, you mean the status of the prime tests all get to 99% and then stop? I wasn't sure what you meant either, but I think I know now.

Yes, that's it. I've noticed it tends to be triggered at times when the client tries to connect but can't due to an interruption at some point down the line (which happens quite frequently with my internet connection). Sometimes it can simply try again when the connection comes back on and all will be OK, but nonetheless it will sometimes freeze.

I've got a theory as to exactly what causes it to freeze: say a test finishes while the internet connection is out. The client moves on to the next test, and tries to report the result. It can't, so it times out and waits before trying again. However, I'm pretty sure that if the *next* test finishes while the previous one's result is still waiting to be reported, *then* it will crash. (Note: this is just a somewhat unscientifically-derived hypothesis; I haven't done any hard tests to see if it's true.)

[gd_barnes]

Halfway through the rally, we have:

9 new primes and 2 confirmed primes.

All have been reported. Nice work everyone. That was actually quite a bit better than expected for the range that we processed. Let's keep 'er going for the 2nd half!

Now I have one question: Where is my prime? 10 quads haven't found one in nearly 3 days or 20K+ tests. Ergh!


Gary

[glennpat]

Originally Posted by gd_barnes

Halfway through the rally, we have:

9 new primes and 2 confirmed primes.

All have been reported. Nice work everyone. That was actually quite a bit better than expected for the range that we processed. Let's keep 'er going for the 2nd half!

Now I have one question: Where is my prime? 10 quads haven't found one in nearly 3 days or 20K+ tests. Ergh!


Gary

I must of got the prime headed for you since you we're do. Looks like I was really lucky.

Nice going all you teams that are ahead of XtremeSystems (which is all of them ) .

[gamer007]

At least you guys are finding primes. The only prime I found was one hours before the rally started.

[PCZ]

Same here.
I think i must be a composite proving specialist :)

PS
Output will be slightly down for a couple of hours, got a few boxes off line.
Cant do much about it as they are remote.