[QUOTE=Prime95;119569]I'm getting P-1 assignments ok. Do the server web pages show the computer as having the correct work preferences? Can you send me the prime.log file?[/QUOTE]

I set the instance back to TF-LMH so I don't know if the prime.log will be much good. I will reset back to P-1 again. I have noticed that for some reason my RAM allocations have reset to 8MB (I'm not certain if I did that and forgot or if it is a bug). Would that have an effect on whether P-1 TF work would be accepted? I didn't get any warnings when I chose it.

I didn't check the server pages, to load test I have the maxeponents rule set to 1200 on a couple of the instances and as a result the assignement page doesn't work that well. It's extremely slow and it still hasn't cleared the Double-Check my instance inadvertently grabbed despite me having deleted it from my worktodo some time ago (I have unreserved it about 5 times). I mailed Scott about that and he said he would have a look (not sure if he got round to it though, I'm sure he has plenty to do!).

I will re-try and monitor to see if the same problem occurs. Have a good weekend folks. a Beta tester toast: May your bugs be plenty and May George's bugspray be mighty :razz:

:party:

[QUOTE=harlee;119618]Anything wrong with the Server? Just got home and I can't get any work and I can't log into my account (Squeeky_Squirrel).[/QUOTE]

I just checked and everything seems to working now - getting work and am able to log onto the server.

[QUOTE=harlee;119621]I just checked and everything seems to working now - getting work and am able to log onto the server.[/QUOTE]

I've been doing a fair number of multi-million record updates that take about 20 minutes each.

[QUOTE=Bundu;119552]Every once in a while I get this

[Nov 29 23:05] pnErrorDetail=Update t_users failed, user_id: 40[/QUOTE]

I've made a SQL change. Let me know if this happens again.

I was running some LMH-TF exponents to 2^58 and 2^60 bits. Noticed that I found a factor in the 59-bit range but it was only worth 0.0001 GHz-days where not finding a factor to 2^58 bits was worth 0.0002 GHz-days.

[CODE][Nov 30 20:26] Trial factoring of M186736943 to 2^58 is 0.12% complete.
[Nov 30 20:26] Trial factoring of M186736943 to 2^59 is 0.06% complete.
[Nov 30 20:26] M186736943 has a factor: 502941880213141361[/CODE]

P4_2600_v255 186736943 F 2007-12-01 01:26 0.0 502941880213141361 0.0001

P4_2600_v255 186735331 NF 2007-12-01 01:03 0.0 no factor to 2^58 0.0002

Server feature requests:

1.) I have set one CPU to run ECM on small mersenne numbers, and the second to factor small fermat numbers. When I view my result details, I see umpteen exponents with one curve finished, and I have to look carefully (and scroll through several pages) to find the few curves which have yielded a factor. It would be nice if there would be an option to view only such results which yielded a factor.

2.) The assignment report gives the assigned exponents, but it doesn't say if a certain exponent x is for 2^x+1 or 2^x-1.

Edit - Bug Report: I just see that the estimated completion date of some of my exponents are 18 days in the past:

[code]
Core2Duo 0 128767 ECM 0.0% 0 0 2007-12-01 22:10 2007-12-02 15:10 2007-12-01 15:10 2007-12-01 15:10
Core2Duo 0 131441 ECM 0.0% 18 -18 2007-11-13 10:13 2007-11-14 02:06 2007-11-13 02:06 2007-11-13 02:06
Core2Duo 0 244603 ECM 0.0% 0 0 2007-12-01 22:10 2007-12-02 14:44 2007-12-01 14:44 2007-12-01 14:44
Core2Duo 0 356387 ECM 0.0% 0 0 2007-12-01 20:32 2007-12-02 12:30 2007-12-01 12:30 2007-12-01 12:30
Core2Duo 0 667423 ECM 0.0% 19 -18 2007-11-13 08:56 2007-11-14 01:25 2007-11-13 01:25 2007-11-12 23:36
Core2Duo 0 667501 ECM 0.0% 18 -18 2007-11-13 10:04 2007-11-14 01:25 2007-11-13 01:25 2007-11-13 01:16
Core2Duo 0 667687 ECM 0.0% 18 -18 2007-11-13 12:12 2007-11-14 04:06 2007-11-13 04:06 2007-11-13 04:06
Core2Duo 0 675239 ECM 0.0% 0 0 2007-12-01 20:21 2007-12-02 12:19 2007-12-01 12:19 2007-12-01 12:19
Core2Duo 0 675449 ECM 0.0% 0 1 2007-12-02 00:20 2007-12-02 17:19 2007-12-01 17:19 2007-12-01 17:19
Core2Duo 0 675457 ECM 0.0% 0 0 2007-12-01 21:00 2007-12-02 12:34 2007-12-01 12:34 2007-12-01 12:34
Core2Duo 0 675481 ECM 0.0% 0 1 2007-12-02 00:46 2007-12-02 17:46 2007-12-01 17:46 2007-12-01 17:46
Core2Duo 0 675553 ECM 0.0% 0 1 2007-12-02 01:13 2007-12-02 18:12 2007-12-01 18:12 2007-12-01 18:12
Core2Duo 0 675559 ECM 0.0% 0 1 2007-12-02 01:39 2007-12-02 18:39 2007-12-01 18:39 2007-12-01 18:39
Core2Duo 0 675617 ECM 0.0% 0 0 2007-12-01 19:26 2007-12-02 12:10 2007-12-01 12:10 2007-12-01 12:10
Core2Duo 0 675617 ECM 0.0% 0 0 2007-12-01 21:27 2007-12-02 13:01 2007-12-01 13:01 2007-12-01 13:01
Core2Duo 0 675713 ECM 0.0% 0 0 2007-12-01 19:54 2007-12-02 12:16 2007-12-01 12:16 2007-12-01 12:16
Core2Duo 0 675713 ECM 0.0% 0 0 2007-12-01 21:59 2007-12-02 13:52 2007-12-01 13:52 2007-12-01 13:52
Core2Duo 0 1279303 ECM 0.0% 19 -18 2007-11-13 07:03 2007-11-14 01:25 2007-11-13 01:25 2007-11-12 23:13
Core2Duo 0 1279307 ECM 0.0% 19 -18 2007-11-13 07:49 2007-11-14 01:25 2007-11-13 01:25 2007-11-12 23:17
Core2Duo 0 1279459 ECM 0.0% 19 -18 2007-11-13 08:34 2007-11-14 01:25 2007-11-13 01:25 2007-11-12 23:31
Core2Duo 0 1279583 ECM 0.0% 18 -18 2007-11-13 09:42 2007-11-14 01:25 2007-11-13 01:25 2007-11-13 00:26
Core2Duo 0 1279757 ECM 0.0% 18 -18 2007-11-13 10:56 2007-11-14 02:25 2007-11-13 02:25 2007-11-13 02:25
Core2Duo 0 1279853 ECM 0.0% 18 -18 2007-11-13 11:46 2007-11-14 03:15 2007-11-13 03:15 2007-11-13 03:15
Core2Duo 0 1280107 ECM 0.0% 18 -18 2007-11-13 13:02 2007-11-14 04:55 2007-11-13 04:55 2007-11-13 04:55
Core2Duo 0 1280129 ECM 0.0% 18 -18 2007-11-13 13:52 2007-11-14 05:45 2007-11-13 05:45 2007-11-13 05:45
Core2Duo 0 1280179 ECM 0.0% 18 -18 2007-11-13 14:35 2007-11-14 06:06 2007-11-13 06:06 2007-11-13 06:06
Core2Duo 0 1280231 ECM 0.0% 18 -18 2007-11-13 15:20 2007-11-14 06:11 2007-11-13 06:11 2007-11-13 06:11
Core2Duo 0 1280333 ECM 0.0% 18 -18 2007-11-13 16:03 2007-11-14 06:30 2007-11-13 06:30 2007-11-13 06:30
Core2Duo 0 1294483 ECM 0.0% 0 0 2007-12-01 23:02 2007-12-02 16:02 2007-12-01 16:02 2007-12-01 16:02
Core2Duo 0 1294823 ECM 0.0% 0 0 2007-12-01 23:53 2007-12-02 16:53 2007-12-01 16:53 2007-12-01 16:53
Core2Duo 1 2097152 ECM C1S2, 43.799999% 15 0 2007-12-01 13:37 2007-12-01 20:06 2007-11-30 20:06 2007-11-16 23:42
Core2Duo 1 2097152 ECM C1S2, 43.799999% 11 0 2007-12-01 20:26 2007-12-01 20:06 2007-11-30 20:06 2007-11-20 17:40
Core2Duo 1 2097152 ECM C1S2, 43.799999% 10 1 2007-12-02 03:14 2007-12-01 20:06 2007-11-30 20:06 2007-11-21 19:12
Core2Duo 1 2097152 ECM C1S2, 43.799999% 9 1 2007-12-02 10:03 2007-12-01 20:06 2007-11-30 20:06 2007-11-22 07:11
Core2Duo 1 2097152 ECM C1S2, 43.799999% 9 1 2007-12-02 16:52 2007-12-01 20:06 2007-11-30 20:06 2007-11-22 10:42
Core2Duo 1 2097152 ECM C1S2, 43.799999% 7 1 2007-12-02 23:41 2007-12-01 20:06 2007-11-30 20:06 2007-11-24 14:22
Core2Duo 1 2097152 ECM C1S2, 43.799999% 5 2 2007-12-03 06:30 2007-12-01 20:06 2007-11-30 20:06 2007-11-26 04:00
Core2Duo 1 2097152 ECM C1S2, 43.799999% 5 2 2007-12-03 13:19 2007-12-01 20:06 2007-11-30 20:06 2007-11-26 14:23
Core2Duo 1 2097152 ECM C1S2, 43.799999% 5 2 2007-12-03 20:08 2007-12-01 20:06 2007-11-30 20:06 2007-11-26 14:23
Core2Duo 1 2097152 ECM C1S2, 43.799999% 5 3 2007-12-04 02:56 2007-12-01 20:06 2007-11-30 20:06 2007-11-26 14:23
Core2Duo 1 2097152 ECM C1S2, 43.799999% 2 3 2007-12-04 09:45 2007-12-01 20:06 2007-11-30 20:06 2007-11-29 10:37
Core2Duo 1 2097152 ECM 0.0% 2 5 2007-12-06 00:08 2007-12-01 20:06 2007-11-30 20:06 2007-11-29 10:37
Core2Duo 1 2097152 ECM 0.0% 0 6 2007-12-07 16:21 2007-12-02 02:13 2007-12-01 02:13 2007-12-01 02:13[/code]

Here is another one - a 60-bit factor is only getting 0.0002 GHz-days

P4_2600_v255 187929143 F 2007-12-01 21:05 0.1 783293698887158111 0.0002

Shouldn't no factor to 2^62 take twice as long as as 2^61bits?

P4_2600_v255 189789283 NF 2007-12-02 21:16 0.1 no factor from 2^62 to 2^63 0.0052
P4_2600_v255 189789283 NF 2007-12-02 21:11 0.1 no factor from 2^61 to 2^62 0.0026
P4_2600_v255 189789283 NF 2007-12-02 21:08 0.1 no factor to 2^61 0.0016

The range is twice as large, but since the primes become more spread out the further you go, you don't necessarily have twice as many factors to test.

[QUOTE=harlee;119758]Shouldn't no factor to 2^62 take twice as long as as 2^61bits? [/QUOTE]

Yes. Your post indicated "no factor to 2^61" which would include credit for 61-bit, 60-bit, 59-bit, etc. factors.

Also, different bit levels are "harder" to factor. The PHP code (timings came from a Core 2 Duo benchmark).

function credit_get_TF_timing( $bits )
{
if ( $bits <= 61 ) return (2.4*0.00465);
elseif ( $bits == 62 || $bits == 63 ) return (2.4*0.00743);
elseif ( $bits == 64 ) return (2.4*0.00711);
elseif ( $bits >= 65 ) return (2.4*0.00707);
}

[QUOTE=harlee;119636]Noticed that I found a factor in the 59-bit range but it was only worth 0.0001 GHz-days where not finding a factor to 2^58 bits was worth 0.0002 GHz-days.[/QUOTE]

Was this due to the fact that finding a factor saved you from testing all the 59-bit factors?