[quote=Mini-Geek;126533]I'm assuming you're referring to hyperthreading. As I recall from Prime95, and LLR is essentially the same AFAIK, it's better to run only one instance (but I could be wrong).
Please post the results of the test you run.[/quote]

[quote=Anonymous;126534]Actually, with hyperthreading, you should definitely not see any performance [I]loss[/I] with running two instances; instead, you will quite likely see a small performance [I]increase[/I]. (Probably, how large or small that increase is will depend on the application you're using.)[/quote]


lol, I have no idea about that technical stuff.

All I know is that when I ran LLR before, I was able to get something like 30-40% more total processing done when I ran 2 instances. I did nothing to the machine settings to get this result. All I can tell you is that's its a 3.2 Ghz Dell desktop from 2004.

As an example, if an LLR test took 100 secs. running one instance, it took ~150 secs. when running 2 instances. So in 300 secs., I was able to get 4 tests running 2 instances and 3 tests running 1 instance or 33% more testing with 2 instances. (Close to the exact test that I remember.)

I'm on my way home shortly...I'll try it first with some manual LLR tests and then with LLRnet tests. I'll let you know how it comes out.


Gary

[quote=AES;126536]I'm seeing a lot of messages like this in the proxy console. What do you make of this?

[code]Result 649/282615 succesfully sent to the server.
Successfully connected to server!
connection request from 7801a8c0:1736 (socket 388)
Proposing pair 465/282619 to PC_3
net_Recv : bad header 'LOGG'
Could not log on to the server
LOGGING OUT
recv error res=-1, errno=0
Could not send result.
The server refused your new result :
either someone else computed it already,
either the server is now configured to
work on other numbers.recv error res=-1, errno=0
connection closed (socket 400)[/code][/quote]

Never seen that before but I am receiving a few (20 or more) rejected pairs from you.
I advise all before the start of the rally to increase the cache size to 10 or 15 because I suppose there will be some delay on client/server connection.

seems there're some server problems?!
carlos is not online! hope he fix it!

[quote=kar_bon;126543]seems there're some server problems?!
carlos is not noline! hope he fix it![/quote]

I will have to reconfigure the proxy back to my other work. hopefully i'll have time to flush it back out and join the rally tomorrow.

[quote=kar_bon;126543]seems there're some server problems?!
carlos is not online! hope he fix it![/quote]
Yes, I'm experiencing problems, too--none of my clients are able to pick up any work or report any results. Fortunately I was able to switch my primary machine back to some manual work I've got sitting around, but Carlos had still better get it fixed in time for the rally...

[quote=Anonymous;126551]Yes, I'm experiencing problems, too--none of my clients are able to pick up any work or report any results. Fortunately I was able to switch my primary machine back to some manual work I've got sitting around, but Carlos had still better get it fixed in time for the rally...[/quote]
Update: I tried troubleshooting what's going on, and it would seem that the port 100 server is down, but not the port 300 one. If I use the nc command in Linux to attempt a connection with the server, I can at least establish a connection on port 300, but I get a "connection refused" for port 100.

yep, so to me: 300 ok, 100 down!
hope carlos won't sleep tooooooo long :grin:

OH!! And I thought I was hallucinating as I tried to test my Athlons and my 3.2-Ghz P4. They both went right into sleep mode

It's 4:30 AM in Portugal now. I'm sure he'll get it fixed in the morning. I'll be up 2 hours before the rally to test them then. Back to sieving...but I'll test again before I go to bed in about 4 hours.

I hope my work machines are able to start again. Unfortunately this is a big loss of processing power for me overnight on 7 machines and I just set them to port 100 this afternoon. Things had been running so well for the last week that I only set them for 5 pairs in the cache. It figures that that would happen after I had 100 pairs in the cache for port 300 all week.

When Carlos gets things working again, can anyone tell me if everyone's machines will automatically start recieving k/n pairs again? Or will LLRnet have to be restarted on each machine?

I may have to make a stop into work 1 hour before the rally.


Gary

[quote=kar_bon;126556]yep, so to me: 300 ok, 100 down!
hope carlos won't sleep tooooooo long :grin:[/quote]

You a major night owl Karsten! 5:30 AM there, right?

yeah, the last nights...
my daughter create a pizza in front of her bed 2 hours ago! :grin:
i had to clean it up! 2 nights ago the other was ill the same way all night!
so what can i do: stay here and wait! wait for the next illness this time online! :grin:

[quote=kar_bon;126560]yeah, the last nights...
my daughter create a pizza in front of her bed 2 hours ago! :grin:
i had to clean it up! 2 nights ago the other was ill the same way all night!
so what can i do: stay here and wait! wait for the next illness this time online! :grin:[/quote]


Ewwwwwe! Gross! Not the kind of pizza I want. Funny thing is my son and I had pizza tonight! :smile:

How old are your girls? I have a 12-year-old daughter and 15-year-old son.


G