LLRnet servers for NPLB

[vaughan]

lol, I like that. Good one MyDogBuster.

Maybe its time to switch the new server to NPLB. My son built a little rig over the last few weeks and put Gentoo 64 on it.

http://mercurylampe.org/dad/

[gd_barnes]

I have asked David to change his ports to a JobMaxTime of 2 days. I hope that will eliminate most or all of the rejected pairs that are happening from time to time.

[vaughan]

The e-mail notification of primes isn't working for me. Dave is trying to investigate.

How do I know?

Yesterday bok's stats showed I had found 137 primes so far. Today its 140 when I wrote to Dave. Now its showing 141. Just checked my e-mail and there are no NPLB messages.

[gd_barnes]

Originally Posted by vaughan

The e-mail notification of primes isn't working for me. Dave is trying to investigate.

How do I know?

Yesterday bok's stats showed I had found 137 primes so far. Today its 140 when I wrote to Dave. Now its showing 141. Just checked my e-mail and there are no NPLB messages.

Bummer. I'm sending you a PM now. ...and after just now checking again, you now have 5 total primes...total should be 142 once everything is updated. Congrats!

[MyDogBuster]

Gary, your ports seem to be malfunctioning. Welcome home.

[gd_barnes]

Originally Posted by MyDogBuster

Gary, your ports seem to be malfunctioning. Welcome home.

Oh holy crap. I got an Email from Max saying he had done something weird and not only is he in bed now, he's heading out of town himself tomorrow I think.

He was trying to get a working version of PRPnet 2.4.1 installed on my server machine and ended up having to install Wine, which I guess allows Linux to execute Window's programs or something like that. The problem was, he was having problems installing Wine because the version of Linux is too old so I will quote what he said he had to do:

Essentially, the system wasn't letting me install Wine (or any new software, for that matter) because Ubuntu discontinued support for version 7.04 a long time ago. As you may recall, crunchford and, I believe, one other AMD (the one right after it, I think) are the only two computers still on 7.04. Anyway, what I had to do to get around this was reconfigure the package manager to point to the software repositories for version 8.04, which is still supported. This allowed me to install the software just fine.

I'm a little nervous about him having to reconfigure the package manager without the OS being upgraded to (at least) Ubuntu 8.04.

The machine is running very cool and connected to a battery backup. I haven't had any problems with having to recycle my router or anything like that. No bad weather. No blips. Nothing. It has to be something with what Max did or just something altogether random.

At this point, there are so many various servers on there, both public and private, I almost hate to shut the machine down and restart them all for fear I'll do something wrong. I'll just try killing the two main NPLB servers and restarting them and see if that does the trick. I'm pretty familiar with what to do on those.

If that doesn't work, I'm hoping that Max sees this and has time to look into it before he leaves on Thursday. I'll also send him an Email. If he can't figure out the problem, we may be without ports G4000 and G7000 (and possibly the one CRUS PRPnet server as well as his and my personal PRPnet servers) for a week or more.

I may be going off the deep end a little here. I hope so.


Gary

[gd_barnes]

Originally Posted by MyDogBuster

Gary, your ports seem to be malfunctioning. Welcome home.

Hum weird. Yep, I went off the deep end. lol

They seem to be working fine. I checked the results and knpairs and they are being updated at a regular interval. I also checked the pairs processed per hour. It dropped in half for the 1 AM and 2 AM CDT hours but seemed to be back to normal for the 3 AM CDT hour.

I'll be on and off for another hour. Let me know if you experience any more problems.

With the battery backup and extra fan keeping the machine extra cool, I fully expect my servers to now be virtually as reliable as David's servers. The only exception is if we threw a huge load at them. We haven't had any rallies since I added those 2 safe guards so I couldn't guarantee their reliability with a big load. But once I build my new server machine and get a fixed IP address, they should be able to handle rallies, if and when we have them again.

As David used to say "It's just one of those things that happens from time to time.".

[mdettweiler]

Hmm...strange. I don't know what could have caused that. I'm not seeing anything amiss. My guess is that it was an IP change or something, which usually serves to take the servers "offline" for many people for up to a few hours.

One thing to note: the crazy stuff I did with the package manager yesterday should NOT, under any circumstances, affect the servers. The only reason I mentioned that to Gary was so that he wouldn't unknowingly try to run the software update tool and blow up the whole system.

The only other tweak I made last night was switching my personal PRPnet server over to 2.4.1 and disabling the automatic results copy-off for that server to sidestep an incompatility with one of my scripts. I did not touch any of the other servers, and a quick look at the results folder shows that those all copied off without issue.

Anyway, looks like everything is all set. If everything goes smoothly until I get back late on the 22nd, then the only problem that might occur is an IP address change, which can be remedied as usual. (Restart DNS clients, etc.)

Boy, now I'm really glad I checked my email this morning before I left!

[MyDogBuster]

My guess is that it was an IP change or something

I think Max is right. I did have to flush dns on 2 machines. The only thing irritating about an IP change is that it takes hours to filter throughout the system. One would think that in this day and age, something like an IP address change should take seconds, and if it doesn't, why not.

[gd_barnes]

Max,

The ETA on the drying of port IB4000 is about 3 days at the current rate of processing. This is just a reminder of the timeframe for setting up the new PRPnet server for it and loading in n=740K-750K. I assume some behind-the-scenes stuff will also be needed to interface the new PRPnet results with the stats pages.

If you can send me a PRPnet client care package for 4 cores, that would be helpful in getting my machines switched over to the new server when the time comes.


Gary

[mdettweiler]

Originally Posted by gd_barnes

Max,

The ETA on the drying of port IB4000 is about 3 days at the current rate of processing. This is just a reminder of the timeframe for setting up the new PRPnet server for it and loading in n=740K-750K. I assume some behind-the-scenes stuff will also be needed to interface the new PRPnet results with the stats pages.

If you can send me a PRPnet client care package for 4 cores, that would be helpful in getting my machines switched over to the new server when the time comes.


Gary

Okay, thanks. Within the next day or two I'll get client packages posted for PRPnet 2.4.6 and get the new server set up. Dave and I are still working on a couple final details for importing PRPnet results into the DB, but that shouldn't be too long and we can go ahead with starting up the server anyway.