Rejected k/n pairs
 

Bryan from RieselSieve project said this about the rejected pairs:

[quote]Rejected is any k/n pair that comes back that is not either in the jobs.txt or jobs table (file vs mysql setup), whether it's too late, the server received the result, but the client did not receive the acknowledgment, the client never had the k/n pair (someone put it in, basically), or any number of reasons. Basically, the server didn't think that user should have had the k/n pair they returned.[/quote]I'm getting a bunch of rejected files from michaf and I think he should be credit for them even if others later receive the pairs to re-test. Meanwhile I increased the job prune to each two days instead of one, maybe this was the problem. Thoughts please.

Carlos

EDIT: I'm going to sleep. Just checked again the file and for the last two hours the server has been receiving all k/n pairs from michaf as rejected ones...

Golden slumbers...

Luigi

Hmm... nothing I did wrong hopefully?
Everything looks ok here, albeit a too low prime content.
As far as I can tell from the lresults.txt file, all results have been sent successfully to the server (manual scan of it, might have missed a few)

And don't bother too much to give the credit, I don't mind being last in the race :)


[QUOTE=em99010pepe;125236]Bryan from RieselSieve project said this about the rejected pairs:

I'm getting a bunch of rejected files from michaf and I think he should be credit for them even if others later receive the pairs to re-test. Meanwhile I increased the job prune to each two days instead of one, maybe this was the problem. Thoughts please.

Carlos

EDIT: I'm going to sleep. Just checked again the file and for the last two hours the server has been receiving all k/n pairs from michaf as rejected ones...[/QUOTE]

I really don't know what's happening. I'll tell Karsten to count the rejected files because they are good, the residues are correct.

michaf, may i ask which os you are running?

My results seem to be going to the rejected file as well as the lresults file. The residues all appear to be correct. I'm seeing strange client behavior in 64-bit Vista, but the results and log files are all ok.

You both are getting rejected pairs...

[quote=AES;125290]michaf, may i ask which os you are running?
[/quote]

I was not receiving rejected pairs when he was running for CRUS project.

I have a desktop running XP-home
and a laptop running Vista Home Premium.
both running on 32bits

edit:
Hey... I just noticed that my laptop AND my desktop were crunching the SAME number:
887*2^328654-1

Any reason as why I get two of the same?

[QUOTE=AES;125290]michaf, may i ask which os you are running?

My results seem to be going to the rejected file as well as the lresults file. The residues all appear to be correct. I'm seeing strange client behavior in 64-bit Vista, but the results and log files are all ok.[/QUOTE]

[QUOTE=michaf;125335]
edit:
Hey... I just noticed that my laptop AND my desktop were crunching the SAME number:
887*2^328654-1

Any reason as why I get two of the same?[/QUOTE]

look up your llr-clientconfig.txt on both:
get sure they different services and guiport adresses.

compare the workfile.txt's! equal?

all 1136 rejected k/n-pairs i got from Carlos are double work of the same user.
seems that 2 clients did same pairs and sent their results, but 2nd was rejected.

1136 pairs:
346 from AES
10 from paratima
780 from michaf

attachment contains table with:
- first 9 columns contains data from rejected file
- next comment and index from result-file
- next 8 columns data from result-file

Karsten

I think I'm the dumb one here...
I made four dir's accessible from both my machines, each containing a llr-net
I intended to start each core in a different dir, but I apparently ended up using the same dir's twice.:shock:

If this was the problem, I'll be sending no double results from now on...

Ah well... at least we have some double-checks done then...

Karsten,

What about adding a column at the score table for double-check pairs?

michaf,

I already stop receiving rejected pairs. I think we all learned by your mistake.

Carlos