PRPnet 2nd drive-51 bases with <= 5 k's to n=250K

[gd_barnes]

Lennart,

It looks like the same problem is happening again. You have found that:

10968*61^102738-1 is prime!

(Actually I'm sure that you've found it to be PRP and cannot prove it to be prime.)

This test has hung on one of your clients for several hours, probably because it cannot prove a PRP. I proved it prime on my machine after seeing the hung test.

The problem is that k=10968 is continuing to be handed out by the server.

Can you please check that all of your clients can prove PRPs and return this test to the server?


Thank you,
Gary

[gd_barnes]

Originally Posted by gd_barnes

Lennart,

It looks like the same problem is happening again. You have found that:

10968*61^102738-1 is prime!

We have what appears to be yet another hung client with a PRP that has been sitting for several hours:

242*67^105312-1 is prime!

Both of the recent primes were on client 14 and were likely found PRP but not returned to the server. Once again, I proved it on my machine.

If I happen to be online, I am now checking any tests that are hung in the queue for a period that is longer than that person's normal testing return time.

[Lennart]

It is the same issue on that computer. They are in resultlog as PRP but close the client when starting N+1 test.
I have changed pfgw to pfgw64s 3.4.9 and restarted the client.Now it is doing the N+1 test, I hope it will be reported to server also.

Lennart

[Lennart]

10968*61^102738-1 is Prime

242*67^105312-1 is Prime

Lennart

[rogue]

Lennart, can you be more specific by what you mean by "hung"? Are you saying that the PRPNet client is stuck due to unexpected results from pfgw or that pfgw is stuck when trying to do the primality tests?

[gd_barnes]

Originally Posted by rogue

Lennart, can you be more specific by what you mean by "hung"? Are you saying that the PRPNet client is stuck due to unexpected results from pfgw or that pfgw is stuck when trying to do the primality tests?

I'm the one that used the word "hung".

The clients get "stuck" when PFGW is trying to do the primality test. This is evidenced by the fact that the test sits in the port 1400 queue (on the web page in the 1st post of this thread) for hours waiting for the client to return the result to the server.

Is this something that can be fixed in PRPnet? I'm thinking that if the client cannot prove a PRP, it should still return it to the server without "hanging up" or "closing".

[Lennart]

Originally Posted by rogue

Lennart, can you be more specific by what you mean by "hung"? Are you saying that the PRPNet client is stuck due to unexpected results from pfgw or that pfgw is stuck when trying to do the primality tests?

pfgw does the prp test and write it to the log. When it start the primalty proof the clint close.

After that I changed to pfgw64s 3.4.9 and restarted the client pfgw did the primalty test and reported it back to the server

First I was using pfgw64 3.4.5

Lennart

[rogue]

Originally Posted by Lennart

pfgw does the prp test and write it to the log. When it starts the primalty proof the client closes.

I believe that what is happening is that the client is thinking that Lennart terminated PFGW with ^C. A message would be in the log if that were happening. Presuming that is the case, the client will shut down. The problem is that the output from PFGW is "incomplete", i.e. it is missing something that tells it that PFGW completed its processing before stopping.

I don't know if there is much I can do about this in the PRPNet client because it is in essence stating that PFGW was terminated, thus it should terminate. I find it interesting that the PRP test did not require glibc, but the primality test does. I could probably put something into PFGW to make it require glibc for any test as that would have helped us discover this problem earlier.

[Lennart]

[2011-08-06 23:21:22 CEST] crus: INFO: Test for 1520*61^102587-1 was accepted
[2011-08-06 23:21:22 CEST] crus: INFO: All 1 test results were accepted
[2011-08-06 23:21:23 CEST] crus: Getting work from server nplb-gb1.no-ip.org at port 1400
[2011-08-06 23:21:27 CEST] crus: PRPNet server is version 4.3.5
[2011-08-06 23:29:19 CEST] crus: Test results not found in file [work_crus.out]. Assuming user stopped with ^C
[2011-08-06 23:29:19 CEST] Total Time:127:44:57 Total Tests: 945 Total PRPs Found: 0
[2011-08-06 23:29:19 CEST] Client shutdown complete
[2011-08-07 12:36:25 CEST] PRPNet Client application v4.3.5 started
[2011-08-07 12:36:25 CEST] User name sm5ymt at email address is sm5ymt@pekhult.se
[2011-08-07 13:07:53 CEST] 10968*61^102738-1 proven prime by pfgw64s. Time: 1888 seconds
[2011-08-07 13:07:53 CEST] crus: 10968*61^102738-1 is prime
[2011-08-07 13:07:53 CEST] Total Time: 0:31:29 Total Tests: 1 Total PRPs Found: 1
[2011-08-07 13:07:54 CEST] crus: Returning work to server nplb-gb1.no-ip.org at port 1400
[2011-08-07 13:07:55 CEST] crus: INFO: Test for 10968*61^102738-1 was accepted
[2011-08-07 13:07:55 CEST] crus: INFO: All 1 test results were accepted
[2011-08-07 13:07:56 CEST] crus: Getting work from server nplb-gb1.no-ip.org at port 1400

[Lennart]

304*135^114227+1 is Prime

Lennart

[Lennart]

58*200^102363-1 is Prime

Lennart