issue with 2 - P4 machines and Linux Mint 10

Questions and Answers : Unix/Linux : issue with 2 - P4 machines and Linux Mint 10

Author	Message
Brian R. Sampson Send message Joined: 21 Feb 11 Posts: 6 Credit: 281,955 RAC: 357	Message 1080884 - Posted: 24 Feb 2011, 6:54:28 UTC
	Hi All .. my first attempt at Linux and both seem to have the same issue, I think its an issue, not aware of the linux digestive system, but it seems an AMD 2400+ with an XP os is kickin the pants off of them. i brought all 3 puters online in 3 days, just 4 days ago I'll post as much as i can give you from my account manager or boinc, seti is the only project PC #1 : Linux (4 days old) 2.6.35-22-generic W/ GenuineIntel Intel(R) Pentium(R) 4 CPU 3.20GHz [Family 15 Model 4 Stepping 10] (2 processors) BOINC version 6.10.58 4 tasks running and 4 waiting to be validated - credit pending PC #2 Linux (3 days old) 2.6.35-22-generic W/ GenuineIntel Intel(R) Pentium(R) 4 CPU 3.00GHz [Family 15 Model 4 Stepping 1] (1 processors) BOINC version 6.10.58 2 tasks running, 1 validated and 4 waiting to be validated Total credit 17 and then what i am basing my assumption on (2 days old) AuthenticAMD AMD Athlon(tm) XP 2400+ [Family 6 Model 8 Stepping 1] (1 processors) XP Home x86 Edition, BOINC version 6.10.58 Total credit 241 - 1 task running and many validated which one is doing the best work for the cause?, is somthing broke? is there a needed fix? Thanks in advance for bearing with me. megabea45786
	ID: 1080884 ·

Volunteer tester Send message Joined: 9 Apr 02 Posts: 11987 Credit: 18,169,433 RAC: 56,165	Message 1080991 - Posted: 24 Feb 2011, 18:36:59 UTC - in response to Message 1080884.
	I'm not sure what your question is. They all seem to be working. which one is doing the best work for the cause?, Best for credit per watt? Best for highest RAC potential? All computers are welcome to crunch. Are you looking to turn off one of the machines and want to turn off the slowest one? is somthing broke? is there a needed fix? Not as far as I can tell.
	ID: 1080991 ·

Brian R. Sampson Send message Joined: 21 Feb 11 Posts: 6 Credit: 281,955 RAC: 357	Message 1081010 - Posted: 24 Feb 2011, 21:18:13 UTC - in response to Message 1080991.
	ok, well thats good if it's not broke, i wont' fix it, i just didnt understand the big spread is credits, the last machine to come online was leading by a long way 241 to 0 and 241 to 17 when posted, I just found it strange the smaller processor is way out front with credits. Thanks for looking/replying megabea45786
	ID: 1081010 ·

Volunteer tester Send message Joined: 9 Apr 02 Posts: 11987 Credit: 18,169,433 RAC: 56,165	Message 1081020 - Posted: 24 Feb 2011, 22:03:27 UTC - in response to Message 1081010.
	Ah, I see what your question is now. I'd say give it some time. The slower computer is gettig it's credit validated faster because it's so slow - it doesn't have to wait on the quorum partner to return the second copy, because it is always the second copy. To explain further, each workunit is sent out to at least two computers to be crunched. Both copies must be returned so the validator can verify the accuracy of the results. If you always return your copy first, then you'll always be waiting on the other computer, a.k.a. your "quorum partner". If you're slower, you will probably always return your copy after the first person already did, thus it will appear that the credit gets granted faster. Give it 5 weeks and you'll see the differences more clearly.
	ID: 1081020 ·

Brian R. Sampson Send message Joined: 21 Feb 11 Posts: 6 Credit: 281,955 RAC: 357	Message 1081023 - Posted: 24 Feb 2011, 22:10:00 UTC - in response to Message 1081020.
	That all make perfect sense now...fab. explanation...i have another linux machine i hpe to bring online later 2night thank you sir for your time, i will now close this thread Megabea 45786
	ID: 1081023 ·

Brian R. Sampson Send message Joined: 21 Feb 11 Posts: 6 Credit: 281,955 RAC: 357	Message 1081271 - Posted: 25 Feb 2011, 19:10:30 UTC - in response to Message 1081023.
	it's only been 20+ hours and it has already turned around Thanks again megabea45786
	ID: 1081271 ·

skildude Send message Joined: 4 Oct 00 Posts: 8944 Credit: 36,784,860 RAC: 60,956	Message 1081287 - Posted: 25 Feb 2011, 19:58:21 UTC - in response to Message 1081271.
	the only real way of determining which is producing more is the RAC for each box. I have noticed that Linux runs a bit slower than windows. I'm not sure why that is. I would expect the intel boxs to produce a bit more than the AMD. the intel chips can run SSE3 which should run much faster than the SSE on the 2400+ Try using optimized apps and see how much faster you can get all 3 ____________ Proud member of TSWB. End terrorism by building a school
	ID: 1081287 ·

Questions and Answers : Unix/Linux : issue with 2 - P4 machines and Linux Mint 10