No Exponents returned for TF to Low Limits

gjmccrac · 2012-01-09, 14:52

I have a machine the is doing TF-LMH (low limits).

It has been attempting to get new assignments and gets the following error message back.

Code:

No data record of requested type

It seems that once the TF-LMH wave front went into the 600M range it is not working anymore. The recent results report does not seem to show anything being reported for the 600M range.

Grant.

Rodrigo · 2012-01-09, 16:27

I'm having a similar problem with two machines doing TF. For the past 24 hours (maybe longer), every 70 minutes when they go to obtain a new assignment, they've been getting a "CURL library error: Operation timed out after 180 seconds with 0 bytes received".

The highest exponent each machine has is at the 581xxxxxx level, so this jibes with gjmccrac's experience.

Rodrigo

petrw1 · 2012-01-09, 17:15

Only one person seems to be working on and/or getting assignments in the 600M/800M ranges. LingUaan.

He does NOT have them all reserved.

Note: I am not suggesting or accusing anyone of anything; just stating observations.

Dubslow · 2012-01-09, 19:50

Quote:

Originally Posted by Rodrigo

I'm having a similar problem with two machines doing TF. For the past 24 hours (maybe longer), every 70 minutes when they go to obtain a new assignment, they've been getting a "CURL library error: Operation timed out after 180 seconds with 0 bytes received".

The highest exponent each machine has is at the 581xxxxxx level, so this jibes with gjmccrac's experience.

Rodrigo

I experienced something similar, once, yesterday. It worked fine an hour afterwards.

Quote:

Originally Posted by gjmccrac

I have a machine the is doing TF-LMH (low limits).

It has been attempting to get new assignments and gets the following error message back.

Code:

No data record of requested type

It seems that once the TF-LMH wave front went into the 600M range it is not working anymore. The recent results report does not seem to show anything being reported for the 600M range.

Grant.

This is definitely something wrong on PrimeNet's end. Perhaps try manually reserving some assignments? Both in the same range and in other ranges and see if that works.

gjmccrac · 2012-01-09, 20:37

Quote:

Originally Posted by Dubslow

I experienced something similar, once, yesterday. It worked fine an hour afterwards.

This is definitely something wrong on PrimeNet's end. Perhaps try manually reserving some assignments? Both in the same range and in other ranges and see if that works.

I did manually assign exponents in order to fill the work queue on the machine. I believe that there is a Primenet issue with TF-LMH.

Rodrigo · 2012-01-09, 22:32

Neither TF machine is still receiving any new assignments this afternoon.

None of my systems is having problems reporting results.

Rodrigo

diamonddave · 2012-01-09, 22:40

Maybe the constant

Code:

#define MAX_PRIME_SSE2	596000000L	/* SSE2 bit limit */

Is used and shouldn't?

Anyone has an old version of Prime95 (pre SSE2) an could test if the new version introduced the problem?

Couldn't find a version pre 25. Version 25 didn't work either

petrw1 · 2012-01-09, 23:00

Quote:

Originally Posted by diamonddave

Maybe the constant

Code:

#define MAX_PRIME_SSE2	596000000L	/* SSE2 bit limit */

Is used and shouldn't?

Anyone has an old version of Prime95 (pre SSE2) an could test if the new version introduced the problem?

Couldn't find a version pre 25. Version 25 didn't work either

All I could add is that when TF-LMH completed 64 bits about 16 months ago it had no problem going all the way to 999M....well actually I can't for sure recall if this higher range wasn't done manually by CHALSALL, MONST, LINDEx, et al.

Dubslow · 2012-01-09, 23:09

Quote:

Originally Posted by diamonddave

Maybe the constant

Code:

#define MAX_PRIME_SSE2	596000000L	/* SSE2 bit limit */

Is used and shouldn't?

Anyone has an old version of Prime95 (pre SSE2) an could test if the new version introduced the problem?

Couldn't find a version pre 25. Version 25 didn't work either

That's the limit for FFT size, as used for P-1 and LL. TF does not use FFTs, and in fact TF by Prime95 has been done on >999M exponents (with varying to little success). Either way, that's not the issue. PrimeNet refuses to assign work, not Prime95 refuses to do work.

diamonddave · 2012-01-10, 00:43

Quote:

Originally Posted by petrw1

All I could add is that when TF-LMH completed 64 bits about 16 months ago it had no problem going all the way to 999M....well actually I can't for sure recall if this higher range wasn't done manually by CHALSALL, MONST, LINDEx, et al.

I could find an old v24 version and it didn't have TF-LMH worktype

petrw1 · 2012-01-10, 01:42

I think there is more to it...

If there was something stopping TF-LMH from handing out over 599M then it would be handing out the next bit level back at 100M.

But there are none there either.

2012-01-09, 14:52	#1
gjmccrac Aug 2009 Ontario, Canada 2·5·13 Posts	No Exponents returned for TF to Low Limits I have a machine the is doing TF-LMH (low limits). It has been attempting to get new assignments and gets the following error message back. Code: No data record of requested type It seems that once the TF-LMH wave front went into the 600M range it is not working anymore. The recent results report does not seem to show anything being reported for the 600M range. Grant.

2012-01-09, 17:15	#3
petrw1 1976 Toyota Corona years forever! "Wayne" Nov 2006 Saskatchewan, Canada 43×107 Posts	Only one person seems to be working on and/or getting assignments in the 600M/800M ranges. LingUaan. He does NOT have them all reserved. Note: I am not suggesting or accusing anyone of anything; just stating observations. Last fiddled with by petrw1 on 2012-01-09 at 17:16

2012-01-09, 22:40	#7
diamonddave Feb 2004 2⁵·5 Posts	Maybe the constant Code: #define MAX_PRIME_SSE2 596000000L /* SSE2 bit limit / Is used and shouldn't? Anyone has an old version of Prime95 (pre SSE2) an could test if the new version introduced the problem? Couldn't find a version pre 25. Version 25 didn't work either Last fiddled with by diamonddave on 2012-01-09 at 22:56*

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
ERROR: cudaGetLastError() returned 4: unspecified launch failure	MacFactor	GPU Computing	0	2017-12-22 16:04
Error returned on P95 submission	kladner	PrimeNet	4	2017-09-30 19:49
P-1 limits?	nucleon	Hardware	8	2015-04-25 23:01
ERROR: cudaGetLastError() returned 30: unknown error	garo	GPU Computing	10	2013-04-27 19:01
GenefX64 limits	siegert81	Miscellaneous Math	2	2011-02-17 13:37

2012-01-09, 16:27	#2
Rodrigo Jun 2010 Pennsylvania 2·467 Posts	I'm having a similar problem with two machines doing TF. For the past 24 hours (maybe longer), every 70 minutes when they go to obtain a new assignment, they've been getting a "CURL library error: Operation timed out after 180 seconds with 0 bytes received". The highest exponent each machine has is at the 581xxxxxx level, so this jibes with gjmccrac's experience. Rodrigo

2012-01-09, 22:32	#6
Rodrigo Jun 2010 Pennsylvania 1646₈ Posts	Neither TF machine is still receiving any new assignments this afternoon. None of my systems is having problems reporting results. Rodrigo

2012-01-10, 01:42	#11
petrw1 1976 Toyota Corona years forever! "Wayne" Nov 2006 Saskatchewan, Canada 43·107 Posts	I think there is more to it... If there was something stopping TF-LMH from handing out over 599M then it would be handing out the next bit level back at 100M. But there are none there either.