Assignment died of dysentery

[chalsall]

Originally Posted by kladner

Ah Sh!t! I misunderstood your response, and ditched the assignment. My bad.

No problem. Just took two simple SQL statements to get things back.

[kladner]

Will this be under a different key? How should I get the assignment back into my line up?
Thanks for covering for my mistake!

[chalsall]

Originally Posted by kladner

Will this be under a different key? How should I get the assignment back into my line up? Thanks for covering for my mistake!

Damn... The candidate was lost during the transition to AnonSpidy.

Sorry about that. Oh well, at least you hadn't done any work on it....

[kladner]

As you say, it's not a problem. It was at the end of the queue anyway.

Thanks for your efforts, chalsall.

[ckdo]

These are the only 6 entries below 30M in my TF worktodo file:

Factor=N/A,27300059,67,69
Factor=N/A,28260473,68,69
Factor=N/A,28333939,68,69
Factor=N/A,29014831,68,69
Factor=N/A,29014871,68,69
Factor=N/A,29016089,68,69

They're also listed on http://www.gpu72.com/account/assignments/:

27300059    DC TF    67    69    2012-05-30 05:52:48    3.28
28260473    DC TF    68    69    2012-06-01 17:08:16    2.11
28333939    DC TF    68    69    2012-06-01 17:08:16    2.10
29014831    DC TF    68    69    2012-06-01 17:08:16    2.06
29014871    DC TF    68    69    2012-06-01 17:08:16    2.06
29016089    DC TF    68    69    2012-06-01 17:08:16    2.06

Problem: Only one of them is actually assigned to GPU72 - the rest have are assigned for DC to other people, some since before GPU72 actually existed:

27300059    Assigned    Trial factoring to "GPU Factoring" on 2012-05-30
28260473    Assigned    Double-checking to "ANONYMOUS" on 2011-07-05
28333939    Assigned    Double-checking to "ANONYMOUS" on 2011-09-10
29014831    Assigned    Double-checking to "ANONYMOUS" on 2011-12-22
29014871    Assigned    Double-checking to "ANONYMOUS" on 2011-12-22
29016089    Assigned    Double-checking to "ANONYMOUS" on 2012-05-11

For an added twist, both http://www.gpu72.com/reports/estimated_completion/ and http://www.gpu72.com/reports/available/ agree that GPU72 doesn't hold any assignments in the 28M and 29M ranges, respectively.

So... why is GPU72 poaching assignments and distributing them to (un?)suspecting participants?

[chalsall]

Originally Posted by ckdo

So... why is GPU72 poaching assignments and distributing them to (un?)suspecting participants?

Hmmmm... Not sure.

I have a new spider focusing on trying to get old anonymous assignments when they expire. Perhaps there's a bug lurking in there. I'll have to look at its logs.

The reason they don't show up on the estimated completion or available assignments pages is the system has detected that they're not owned by us, and so they're set to be "returned" as soon as the work is completed or they are unassigned/expire.

If you unreserve those five you shouldn't have been given, they'll be out of the system and won't be assigned to anyone else.

[ckdo]

New problem.

From my GPU72 assignments page:

44292841    Lucas-Lehmer    72    --    2012-06-15 15:12:51    71.08
44356061    Lucas-Lehmer    74    --    2012-06-15 15:12:51    71.19
44457869    Lucas-Lehmer    74    --    2012-06-15 15:12:51    71.35

From the exponent status pages:

44292841    Assigned    LL testing to "GPU Factoring" on 2012-06-15
44356061    Assigned    LL testing to "GPU Factoring" on 2012-06-15
44457869    Assigned    LL testing to "GPU Factoring" on 2012-06-15

From my prime.log:

Sending expected completion date for M44292841: Nov 20 2012
PrimeNet error 40: No assignment
ap: Assignment key 8DA85B...9C8D3C belongs to another user

Sending expected completion date for M44356061: Nov 20 2012
PrimeNet error 40: No assignment
ap: Assignment key 5FC64B...9C332C belongs to another user

Sending expected completion date for M44457869: Dec 05 2012
PrimeNet error 40: No assignment
ap: Assignment key 235BE7...287B12 belongs to another user

Lost in transition?

(Unreserved the 28M/29M ones, BTW. Commented out the 44M ones, for now.)

[chalsall]

Originally Posted by ckdo

Lost in transition?

Stupid human error...

The transition to AnonSpidy for candidates below 45M is done manually. Unfortunately, I transitioned from Spidy to Spidy, instead of to AnonSpidy.

I've done the transition properly, updated the DB, and have PMed you the correct AIDs.

[ckdo]

The new AIDs work fine; thanks.

[bcp19]

Originally Posted by chalsall

Stupid human error...

The transition to AnonSpidy for candidates below 45M is done manually. Unfortunately, I transitioned from Spidy to Spidy, instead of to AnonSpidy.

I've done the transition properly, updated the DB, and have PMed you the correct AIDs.

Ya know chalsall, you'd halve your workload if it weren't for you.

[kladner]

I guess this is as good a place as any to note that a DC I had in progress was finished by

25838287No factors below2^69 
P-1B1=285000, B2=3633750 
Verified LL232B4614879A8803 by "Anthony Lieuallen" 
Verified LL232B4614879A8803 by "jg5" on 2012-08-28 
AssignedDouble-checking to "ktony" on 2012-08-21 
Historyno factor for M25838287 from 2^67 to 2^68 [mfaktc 0.19 barrett76_mul32] by "GIMPS Visualization" on 2012-08-21 
Historyno factor for M25838287 from 2^68 to 2^69 [mfaktc 0.19 barrett76_mul32] by "Carsten Kossendey" on 2012-08-21 
History232B4614879A88__ by "jg5" on 2012-08-28

I'm not sure what motivates someone to poach DCs.

EDIT: My thanks again to chalsall for the refinements on the Assignments page of GPU72. The warning of a completed assignment at least allowed me to cut my losses.