Team drive #1: k=400-1001 n=333.2K-600K

[gd_barnes]

Originally Posted by gd_barnes

Oh, didn't expect to hear from you again. Hold off on the results until after you get back unless you see another post here before I go to bed. Then if you have time, feel free to process it before you go but don't knock yourself out doing so.

As it turns out, there are a # of candidates remaining in port 400 around n=462.3K-.6K that were obviously returned to the server by someone who left them in their queue during the rally. I was being fed n>500K on Weds. by that server so I switched to port 8500 on Thurs. but I've now switched back to port 400 to clear them out since the stragglers must have been returned in between than and now.

I have 5 high-speed cores processing the stragglers now. If I start getting fed n>500K candidates tonight before bed, then those stragglers should be cleared out and you can process the results. I'll let you know.


Gary

Anon,

Apparently I have cleared out all of the stragglers in port 400 because I am now getting n=500.7K candidates on that port. There were about 150 or so of them. :surprised

If you have time before you leave, you can process the results for n=458K-463K for Ironbits port 400 server. BUT...if you find a bunch of missing results, say, perhaps 25 or more, around the n=462.4-.6K range, you don't need to process them to me. It's likely that they are in somebody's queue somewhere and just haven't been returned to the server because they haven't reached 7 days yet. If that happens, I'll keep an eye on it and process them as they show up in the server.


Gary

[mdettweiler]

Originally Posted by gd_barnes

Anon,

Apparently I have cleared out all of the stragglers in port 400 because I am now getting n=500.7K candidates on that port. There were about 150 or so of them. :surprised

If you have time before you leave, you can process the results for n=458K-463K for Ironbits port 400 server. BUT...if you find a bunch of missing results, say, perhaps 25 or more, around the n=462.4-.6K range, you don't need to process them to me. It's likely that they are in somebody's queue somewhere and just haven't been returned to the server because they haven't reached 7 days yet. If that happens, I'll keep an eye on it and process them as they show up in the server.


Gary

Okay, I'll try to get them processed as soon as I get the chance.

[mdettweiler]

Originally Posted by Anonymous

Okay, I'll try to get them processed as soon as I get the chance.

Well, it looks like we've got a bunch of "holes"--even though the server shows a minimum n>500K. My guess is that all the "holes" might have been filled in after the daily results file processing run (which, due to IronBits' tweaking, apparently happened a little early, at 4:14 AM his time instead of 7 AM), and thus won't show up until tomorrow's results file.

Oh well, I'll get it when I get back. And I'll still send Adam a completely up-to-date CSV stats import file before I go. Stay tuned...

[Beyond]

486-496 completed ---- results emailed.

the bad news had 2 more machines fail today, makes a total of 8 machines dead now .... the storm last week has taking a toll on my farm.

I'm going to drop the 510-520 range until I can do a complete check and make repairs.

[gd_barnes]

OK, Beyond's issue cements my decision to do just this:

Beginning early Weds. afternoon CDT, I'm stopping all of my other NPLB, CRUS, and twin prime efforts except a few straggling efforts with a few days remaining and putting them on port 400 for at least a week or more. That's nearly 4 more quads for a total of 5 on it and when the straggling efforts finish, I'll have nearly 6 quads on it. It'll be just like the rally before last from my perspective.

We're getting smoked by several projects right now, namely RPS, Riesel Sieve, and 12121. Besides people on vacation, we've encountered a bad primeless range in port 8500 and to a lesser extent port 300. It's time to kick some butt and take no prisoners.

Bruce, just now I've moved my 5 NPLB LLRnet cores from port 8500 to port 300. You're smokin' 8500 and should have it done by late Weds./early Thurs. even without my help. When you're done, if you wouldn't mind switching to port 300, that's where they will be needed next. Thanks for the huge boost there!

Port 300 is currently at n=~482K and has contiguous pairs in it up to n=485K. After that it has pairs from n=496K-500K. After 485K, 496K will be the lowest remaining range in any server so we should just keep right on going with port 300.


Gary

[gd_barnes]

Originally Posted by Beyond

486-496 completed ---- results emailed.

the bad news had 2 more machines fail today, makes a total of 8 machines dead now .... the storm last week has taking a toll on my farm.

I'm going to drop the 510-520 range until I can do a complete check and make repairs.

Would you like us to hold the 510-520 range for you? I could see perhaps holding it up to 3 weeks. The servers have a long way to go to catch up to 510 so no problem there. The only thing I might do is if someone requests a manual reservation, I might send them the lower end of the range. That seems unlikely for a week or more. Mini-Geek has been the only one regularly requesting manual reservations and to a lesser extent Carlos and occasionally Anon when he goes on a trip. If Mini-Geek asks for a range before you're ready to process it, I might slice off 510-511 for him as he likes 1K ranges. Would that be OK? If so, I'll just send you a slightly smaller new file when you're ready to process it.

Sorry to hear about the damage to your machines. I would be livid. You seem to be taking it in stride.


Gary

[Beyond]

I did not mean to release the range, only that I am stopping the tests until I can get everything back to normal.

It seems that most of the damage to the machines at home has been PSU's and surge suppressors. And now I'm getting random reboots on a couple of machines that were doing some sieving work.

At work I lost 2 servers, the UPS's are damaged as well as the psu, replacements are scheduled to arrive today.


On a different note, what server/port would be a good choice for a long term deployment? I have several P4 machines at one of our field offices that I can Borg tomorrow. The problem is I only get out to this office about once every 3 months or more unless there is a problem.

[gd_barnes]

Originally Posted by Beyond

I did not mean to release the range, only that I am stopping the tests until I can get everything back to normal.

It seems that most of the damage to the machines at home has been PSU's and surge suppressors. And now I'm getting random reboots on a couple of machines that were doing some sieving work.

At work I lost 2 servers, the UPS's are damaged as well as the psu, replacements are scheduled to arrive today.


On a different note, what server/port would be a good choice for a long term deployment? I have several P4 machines at one of our field offices that I can Borg tomorrow. The problem is I only get out to this office about once every 3 months or more unless there is a problem.

Oh, OK. I'll keep the entire range reserved for you. If someone requests a manual range, I'll send them a range higher than what has already been reserved.

For long-term deployment, definitely use port 300 as it is our main permanent server for NPLB drive 1. Even if the n-range in it becomes higher than other servers (it will be quite a while before that happens after Bruce dries port 8500), the rest of us can work on the other ports to catch them up.

BTW, I've never heard the term 'Borg'. I assume it means something like 'scroung off of' or 'bum off of' someone or something. lol


Gary

[IronBits]

Borg - To load up a DC client on a computer that doesn't belong to you.
This includes friend/relatives, unsuspecting neighbors, even computers at work can be borged, however,
to be legit, you have permissions to do so, so you won't get fired and/or give DC a black eye. (basic version)
Just like you used to do at work

[em99010pepe]

Gary needs to know Fozzie....

[Mini-Geek]

485.1-.2 and .6-.7 complete no primes, holding lresults until end.