OFFICIAL "SERVER PROBLEMS" THREAD

[storm5510]

I am having a bit of an oddity. Consider the quote below:

Sending result to server: UID: storm5510/7700_Kaby_Lake, M7776673, completed P-1, B1=375000, B2=8156250,
E=6, Wh8: C9B601F8, AID: 344519F9796F9F776D63796F1959003C

PrimeNet success code with additional info:
Result was not needed. P-1 exponent: 7776673, B1: 375000, B2: 8156250
CPU credit is 0.2651 GHz-days.

This is from one of James Heinrich's sub-projects, Exponents that were poorly P-1 factored, which is located here. A person sets their parameters at the top then clicks on the "Search" button. It will show a list of assignments which can be copied and pasted to another location.

These assignments do not have authorization numbers. To get those, a person must run them through either Prime95 or mprime until all have AID's, if a person wants to run them with another application, like gpuOwl. Since I am using mprime, I leave them there. When that is done, I look at the assignments to see if any have "N/A" where the ID should be. I remove any that I may find. I take it these are already assigned. James clearly states a the top of his page to get AID's and not run them without as this may result in "poaching" as some here call it.

My quoted result above has an AID, yet says the result was not needed. I had five of these consecutively This may have been because of the server issues, but I do not see this likely. James does a data regeneration late every evening. I know this applies to his main project, TF>1000M. Whether it applies to what I am try to do, I do not know.

I started another group of 30 about an hour ago. So far, all of the results were not needed, so I stopped running. All 30 started with AID's. 25 remain. I will just wait as see what happens.

[masser]

See exponent report here.

The B1 you used was too small.

[James Heinrich]

Originally Posted by storm5510

So far, all of the results were not needed, so I stopped running. All 30 started with AID's.

You've been poached.

2012-08-27: B1= 95000, B2=1876250  // Nesowa
2020-07-12: B1=500000, B2=8000000  // Ivan M
2020-07-13: B1=375000, B2=8156250  // storm5510

Seems "Ivan M" decided to work in the same range as you but unfortunately decided to not check that the exponents he's working on were already reserved by you, so he's effectively wasted all your effort.
On the plus side, he did P-1 to slightly-better bounds than you.

I would go to https://www.mersenne.org/workload/ and unreserve any P-1 assignments in that range since they're either already poached or likely to be poached. I would avoid working in that range for the next little while since apparently "Ivan M" is unlikely to respect your reservations.

[storm5510]

Originally Posted by masser

See exponent report here.

The B1 you used was too small.

The B1 mprime used. I do not know of a way to alter that, nor for Prime95.

Originally Posted by James Heinrich

You've been poached.

Seems "Ivan M" decided to work in the same range as you but unfortunately decided to not check that the exponents he's working on were already reserved by you, so he's effectively wasted all your effort.

On the plus side, he did P-1 to slightly-better bounds than you.

I have been ordering by exponent. Ordering by probability is the default. The bounds I got from your probability calculator, (Attached image). There are two sides to that and I have never been really clear on which I should use. Each makes a different suggestion. Running these on gpuOwl allows for a specific set of bounds, if wanted. If not specified, it will use its defaults. I am not sure what it bases these on. Caveat: gpuOwl does not submit results as they are completed. They must be submitted manually as a group. kriesel is working on a Perl script to submit them when completed. I know Primenet can handle groups, to a point. I would rather not send them this way if I can get around it.

My poacher may be running based on exponent size. Smaller first. I think I will try leaving it at the default ordering and start at 10-million. Perhaps, I can avoid any pilfering for a while.

A question I have been wanting to ask for a while: Does your daily data regeneration include this area?

Thank you for the feedback.

[James Heinrich]

Originally Posted by storm5510

Does your daily data regeneration include this area?

That data is rebuilt daily at approximately 0600h UTC.

[Madpoo]

Originally Posted by Prime95

I'm not wise in the way of managing a Windows Server. But Aaron seems to be offline, so I've tried restarting the SQLServer service. Fingers crossed, the problems will go away.

I got super focused on a project today and didn't check email.

Thanks for getting in there and giving it a shove. LOL

I was actually going to reboot it tomorrow after patch Tuesday stuff, so I'll be sure to give it a good going-over.

[petrw1]

Originally Posted by Madpoo

I got super focused on a project today and didn't check email.

Thanks for getting in there and giving it a shove. LOL

I was actually going to reboot it tomorrow after patch Tuesday stuff, so I'll be sure to give it a good going-over.

Post #1889.
Factor result is still missing from My Results, PrimeNet, and Mersenne.ca
for 40385207.

[James Heinrich]

Originally Posted by petrw1

UID: petrw1/Magic_8_Ball, M40385207 has a factor: 6402438646760818548080303 (P-1, B1=1500000, B2=30000000, E=12)
My earlier post shows that it belonged to me. Exponent status still has no record of it. Maybe I just need to be patient?

Something got broken, most likely due to whatever weirdness was going on with SQL. The factor itself got recorded but the process died before it recorded the P-1 effort. And now that there is a known factor, P-1 is no longer needed. That's something George will have to fix manually (inserting the P-1 effort into the exponent log).

[storm5510]

Originally Posted by James Heinrich

That data is rebuilt daily at approximately 0600h UTC.

Thank you!

I have not had any problems. Everything seems to be working well today.

[petrw1]

Originally Posted by petrw1

Post #1889.
Factor result is still missing from My Results, PrimeNet, and Mersenne.ca
for 40385207.

Fixed...thanks

[petrw1]

Msg 13609, Level 16, State 1, Server PRIMENET, Procedure sp_gimps_report_global_cleared_1000_table, Line 46
JSON text is not properly formatted. Unexpected character '.' is found at position 1024.