PrimeNet 5.0 Upgrade

[uigrad]

I just realized that all v4 assignments I've turned in since the switch are still listed in my assignment page! Most of them have a negative value for days to go.

Are we supposed manually release these assignments after they are finished and turned in?

[garo]

Are they TFs? There is a known bug with TFs but LLs and DCs should unreserve.

[petrw1]

Originally Posted by garo

Are they TFs? There is a known bug with TFs but LLs and DCs should unreserve.

I have assignments that are not dropping off either. In my observation TFs that find a factor are dropping off. Those that do NOT are staying on the list. I assume that is on George's return list.

[petrw1]

Originally Posted by ixfd64

I don't know if this has been fixed yet, but v5 will reject certain v4 TF results submitted through Prime95 v25.x. The problem is that v25.x.x sends results every time it finishes factoring a bit depth, but v24.x does not send any results until the assignment is completely finished.

Two weeks ago, I upgraded from v24.14 to v25.7 in the middle of an assignment because the GIMPS website said it was safe to do so. It seems PrimeNet does not accept TF results unless it has known data for the lower bit depths. For example, it won't accept "no factor from 2^68 to 2^69" if the exponent has "no factor to 2^64" at most.

Since I upgraded to v25.x in the middle of the TF assignment, PrimeNet never got the the data for the lower bit depths.

Exponent Status Data  
50324467 No factors below 2^63 
 Assigned LL testing to "petrw1" on 2008-10-27 
 History no factor from 2^68 to 2^69 by "petrw1" on 2008-11-07

Yes the Exponent Status Report does look suspect ... it appears to be missing some bits even though I did start at 64.
I wasn't assigned it as LL but rather TF ... the v5 conversion just labelled it as LL.

[ixfd64]

Originally Posted by garo

Submit on the manual pages.

I tried it last time and it didn't work. I just tried it again and it worked this time.

However, the exponent status page now looks like this:

Exponent  Status           Data  
50378749  No factors below 2^64 
          History          no factor from 2^68 to 2^69 by "Danny Chia" on 2008-11-07
          History          no factor from 2^68 to 2^69 by "Danny Chia" on 2008-11-07

Shouldn't it say "no factor from 2^65 to 2^69" instead? The server may be thinking that no factoring has been done between 64 and 69 bits, which is not the case.

Also, I didn't get any credit for it:

CPU Name       Exponent  Result Type Received         age days Result                      GHz-Days 
Manual testing 50378749  NF          2008-11-07 21:39 0.0      no factor from 2^68 to 2^69 0.0000

Is anyone else having the same problem?

[EPF]

[Fri Nov 07 14:43:03 2008]
UID: EPF/EPF04, M48543991 has a factor: 295691936519281003567


Work Results Details

CPU Name Exponent Result Type Received age days Result GHz-Days
v4_computers 48543991 F-PM1 2008-11-07 13:53 10.7 2.5608


PrimeNet Top Trial Factoring Producers

Rank Member Name GHz-Days Attempts Successes |90day 30day 7 day 1 day
------ -------------------- ------------ -------- --------- |----- ----- ----- -----
100 EPF 75.274 34 0 | 50 27 35 1

Success = 0

[petrw1]

Originally Posted by EPF

[Fri Nov 07 14:43:03 2008]
UID: EPF/EPF04, M48543991 has a factor: 295691936519281003567

I had one of those too but I assumed it was because v4 TF assignments are shown as LL. SO if a factor is found during a LL it is assuming it is a PM1 factor by mistake.

[petrw1]

Originally Posted by garo

Submit on the manual pages.

I did so but based on the output I believe it only gave me credit for the 69 bits part and not the 64, 65, 66, 67 oir 68 bits.

No factor lines found: 3
Processing result: M50324467 no factor from 2^66 to 2^67, Wd1: E30D4EE9
DEBUG SQL: insert_row_array(): INSERT t_stats_work_credit ( user_id, cpu_id, team_id, stats_type, attempts, successes, dt_received, source, CPU_GHz_days ) VALUES( '112', '19595', NULL, 1001, 1, 0, GETDATE(), 10, 0.29698237155935 )DEBUG SQL: table t_stats_work_credit inserted 1 row(s)Result was not needed.  TF on M50324467, sf: 66, ef: 67
CPU credit is 0.29698237155935 GHz-days.
Processing result: M50324467 no factor from 2^67 to 2^68, Wd1: E30D4EE9
Result was not needed.  TF on M50324467, sf: 67, ef: 68
CPU credit is 0.59396474311869 GHz-days.
Processing result: M50324467 no factor from 2^68 to 2^69, Wd1: E30D4EE9
DEBUG SQL: insert_row_array(): INSERT t_stats_work_credit ( user_id, cpu_id, team_id, stats_type, attempts, successes, dt_received, source, CPU_GHz_days ) VALUES( '112', '19595', NULL, 1001, 1, 0, GETDATE(), 10, 1.1879294862374 )DEBUG SQL: table t_stats_work_credit inserted 1 row(s)Result was not needed.  TF on M50324467, sf: 68, ef: 69
CPU credit is 1.1879294862374 GHz-days.
Factors found: 0
P-1 lines found: 0
LL lines found: 0
Mlucas lines found: 0
Glucas lines found: 0
ECM lines found: 0

Correction: I also got the 67 and 68 bit credit. Pretty close to 7/8 of total

[g0ods]

Originally Posted by g0ods

I upgraded a machine to 25.7 and it all seemed fine, but when it contacted the server (after a few days) to get new work, the current work exponent, (a double check), has disappeared from 'My Account | Assignments', though it still shows up under 'Summary' and also under 'CPUs'.

Will the result be rejected by the server as not needed? Or will it be accepted and credited properly?

M.

A little digging in the Exponent Status page revealed this...

23030431	No factors below	2^67
	        P-1	                B1=330000
	        Unverified LL	        AC9F8DB99600C2__ by "francesco"
	        Assigned	        Double-checking to "ANONYMOUS" on 2008-10-27

How can I get the exponent that I am working on assigned back to my account/user ID ?

[starrynte]

Originally Posted by g0ods

A little digging in the Exponent Status page revealed this...

23030431    No factors below    2^67
            P-1                    B1=330000
            Unverified LL            AC9F8DB99600C2__ by "francesco"
            Assigned            Double-checking to "ANONYMOUS" on 2008-10-27

How can I get the exponent that I am working on assigned back to my account/user ID ?

If "anonymous" is you, the only way I know of is risky, and that is: unreserving the exponent but keeping the save files, then manually reserving the exponent again. there is a chance that in the period the exponent was unreserved it could have been reassigned

[Old man PrimeNet]

Assignment transfers from an unlinked v4 acct to a v5 acct are viewed by PrimeNet as poaching. If the v4 acct is not linked the assignments belong to v5 user Anonymous acct. Upgrading to Prime95 25.7 in place, however, normally performs the transfer properly.