Rootfinder instead of linesiever this time

dsouza123 · 2003-09-01, 05:16

On running the nfsnetclient it called/launched/spawned the rootfinder.
Normally it is the linesiever.

Why did it do that ?

Has the linesiever exhausted the search space with the previous root(s) ?

Has the 2^757 - 1 entered a new stage of processing ?

Is this normal ?

I have examined the text file documentation and it mentions the rootfinder running after registration, but doesn't mention it re-occuring.

Does it automatically expire/ redo roots at the start of the month ?

Wacky · 2003-09-01, 11:27

Some good guesses. I like the "first of the month" hypothesis. smile

Actually, you must be in the "Texas" pool (one of five). That pool has been working on subproject 2_757M_2 while everyone else was working on 2_757M_1. Last night, we finished 2_757M_2 and so those sievers have now switched to 2_757M_1.

For those interested, let me explain the difference.
Our objective in sieving is to find "enough" relations. We want to try to be efficient in doing so.
We sieve over a two dimensional space. However, the probability of finding a relation at a particular location is not uniform. It is more like 1/(a * b). For technical reasons, we sieve rectangular regions. Since it is more efficient to search near the axes, we search a wide rectangle for small values of "b" and a narrower one for the larger values of "b". These correspond to the two subprojects, 2_757M_2 (0<b<1M) and 2_757M_1 (1M<b<24M), in this case.

The relations from both subprojects are combined, but it is easier for us to administer things by treating them as if they were different factorizations. In either case, the sievers are designed to automatically get the data and process whatever assignment they are given.

2003-09-01, 05:16	#1
dsouza123 Sep 2002 2×331 Posts	Rootfinder instead of linesiever this time On running the nfsnetclient it called/launched/spawned the rootfinder. Normally it is the linesiever. Why did it do that ? Has the linesiever exhausted the search space with the previous root(s) ? Has the 2^757 - 1 entered a new stage of processing ? Is this normal ? I have examined the text file documentation and it mentions the rootfinder running after registration, but doesn't mention it re-occuring. Does it automatically expire/ redo roots at the start of the month ?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
P95 est time to go seems off	bcp19	Software	1	2012-08-03 22:39
How much time	Unregistered	Information & Answers	4	2008-12-20 21:00
New .dat time?	benjackson	Prime Sierpinski Project	16	2008-07-29 07:26
linesiever RAM	junky	NFSNET Discussion	10	2004-02-27 00:41
P3 TF time	PrimeCruncher	Software	30	2003-12-21 05:26

2003-09-01, 11:27	#2
Wacky Jun 2003 The Texas Hill Country 3²×11² Posts	Some good guesses. I like the "first of the month" hypothesis. smile Actually, you must be in the "Texas" pool (one of five). That pool has been working on subproject 2_757M_2 while everyone else was working on 2_757M_1. Last night, we finished 2_757M_2 and so those sievers have now switched to 2_757M_1. For those interested, let me explain the difference. Our objective in sieving is to find "enough" relations. We want to try to be efficient in doing so. We sieve over a two dimensional space. However, the probability of finding a relation at a particular location is not uniform. It is more like 1/(a * b). For technical reasons, we sieve rectangular regions. Since it is more efficient to search near the axes, we search a wide rectangle for small values of "b" and a narrower one for the larger values of "b". These correspond to the two subprojects, 2_757M_2 (0<b<1M) and 2_757M_1 (1M<b<24M), in this case. The relations from both subprojects are combined, but it is easier for us to administer things by treating them as if they were different factorizations. In either case, the sievers are designed to automatically get the data and process whatever assignment they are given.