Welcome to MilkyWay@home

New Server

Message boards : News : New Server
Message board moderation

To post messages, you must log in.

AuthorMessage
Matthew
Volunteer moderator
Project developer
Project scientist

Send message
Joined: 6 May 09
Posts: 217
Credit: 6,856,375
RAC: 0
Message 51801 - Posted: 2 Dec 2011, 21:24:21 UTC

A new server has been ordered. Because delivery times are slow now (due to flooding in Thailand producing slow delivery times for disks), the estimated ship date is December 29, which means we do not expect to get the new server running until the middle to end of January 2012. We will keep watching the current server for repeats of previous software glitches, and hopefully will keep that server running smoothly for the next two months.

Thanks,
Heidi
ID: 51801 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Matthew
Volunteer moderator
Project developer
Project scientist

Send message
Joined: 6 May 09
Posts: 217
Credit: 6,856,375
RAC: 0
Message 51802 - Posted: 2 Dec 2011, 21:24:47 UTC

I'm going to try to get the specs of the new server and post them here soon.
ID: 51802 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
HassanShebli

Send message
Joined: 2 Oct 10
Posts: 74
Credit: 18,362,557
RAC: 0
Message 51803 - Posted: 2 Dec 2011, 22:25:12 UTC - in response to Message 51802.  

Good news.

Can you please make a different website for MW so users will be able to logon to follow the progress in case of another disaster? Or at least a facebook account?

Hope you update the database to avoid future MELTDOWN.


Thanks
ID: 51803 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Matthew
Volunteer moderator
Project developer
Project scientist

Send message
Joined: 6 May 09
Posts: 217
Credit: 6,856,375
RAC: 0
Message 51804 - Posted: 2 Dec 2011, 23:26:23 UTC

Can you please make a different website for MW so users will be able to logon to follow the progress in case of another disaster?


We're thinking of using the old server for that, if all goes well.
ID: 51804 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Matthew
Volunteer moderator
Project developer
Project scientist

Send message
Joined: 6 May 09
Posts: 217
Credit: 6,856,375
RAC: 0
Message 51805 - Posted: 2 Dec 2011, 23:30:15 UTC
Last modified: 2 Dec 2011, 23:31:44 UTC

NEW SERVER SPECIFICATIONS:


Memory: 64GB Memory (8x8GB), 800MHz 2R LV RDIMMs at Std Volt for 1 Proc, Optimized

Processor: Intel® Xeon® X5647 2.93GHz, 12M Cache, 5.86 GT/s QPI, 4C

Hard Drive Configuration: RAID 5 for H700 or PERC 6/i Controllers

Hard Drives: (4) 500GB 7.2K RPM SATA 3.5" Hot Plug Hard Drive

Note: There are, of course, other specifications, but I think these are the most important ones. :)
ID: 51805 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mark Vang
Avatar

Send message
Joined: 26 Feb 11
Posts: 1
Credit: 261,812
RAC: 0
Message 51806 - Posted: 2 Dec 2011, 23:51:48 UTC - in response to Message 51803.  

Ditto this. I would like to see more projects set up a Facebook page (or G+ if you hate FB). You can update the page from your cell phone so even if the entire computer center goes down, you can still deliver updates.

Don't put all your eggs in one basket.

...and...

Keep up the good work!

Good news.

Can you please make a different website for MW so users will be able to logon to follow the progress in case of another disaster? Or at least a facebook account?

Hope you update the database to avoid future MELTDOWN.


Thanks


BOINC Atlantic Team Blog
BOINC Atlantic Team on Facebook
ID: 51806 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
HassanShebli

Send message
Joined: 2 Oct 10
Posts: 74
Credit: 18,362,557
RAC: 0
Message 51809 - Posted: 3 Dec 2011, 11:48:31 UTC - in response to Message 51806.  

Matthew:

Is it dual or single CPU?
ID: 51809 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Werkstatt

Send message
Joined: 19 Feb 08
Posts: 350
Credit: 141,284,369
RAC: 0
Message 51819 - Posted: 3 Dec 2011, 20:43:44 UTC

ID: 51819 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
HassanShebli

Send message
Joined: 2 Oct 10
Posts: 74
Credit: 18,362,557
RAC: 0
Message 51820 - Posted: 3 Dec 2011, 23:53:41 UTC - in response to Message 51819.  

I know the CPU, but I was asking if the server has dual CPU, or only one CPU ?

For MW it is much better to have more CPU power to accomplish more tasks at once and handle the stress efficiently.


ID: 51820 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Ray_GTI-R
Avatar

Send message
Joined: 5 Nov 10
Posts: 69
Credit: 15,064,831
RAC: 0
Message 51821 - Posted: 4 Dec 2011, 1:37:12 UTC - in response to Message 51820.  
Last modified: 4 Dec 2011, 1:40:31 UTC

So the question posed is:- how many physical PROCESSORS can the server MOTHERBOARD accomodate?

(Which is irrespective of/irrelevant to the project being ummm ... served ... on that kit? Correct me if I've got that wrong.)
ID: 51821 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
peeticek_LubosKrutek

Send message
Joined: 12 Sep 09
Posts: 4
Credit: 20,183,160
RAC: 0
Message 51826 - Posted: 4 Dec 2011, 11:55:22 UTC

Sounds great! Good luck guys and hope that old server would be just for www.

Btw - how looks the typical load of these servers? //CPU/RAM/HDD/LAN

The new one has 8 cores - what about running some Milky units on it? :):):)

thanks

p.

P.S. How many of yours computers in LAB running Milky? // theirs configs?

http://www.boincstats.com/signature/user_4533.gif
ID: 51826 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Len LE/GE

Send message
Joined: 8 Feb 08
Posts: 261
Credit: 104,050,322
RAC: 0
Message 51827 - Posted: 4 Dec 2011, 12:33:34 UTC - in response to Message 51805.  
Last modified: 4 Dec 2011, 12:36:50 UTC


Memory: 64GB Memory (8x8GB), 800MHz 2R LV RDIMMs at Std Volt for 1 Proc, Optimized

Processor: Intel® Xeon® X5647 2.93GHz, 12M Cache, 5.86 GT/s QPI, 4C


From cpu specs:
# of Cores: 4
Max TDP: 130 W
Memory Types: DDR3-800/1066
Max Memory Size: 288 GB
Max CPU Configuration: 2

With a proper PSU and a decent board there should be room for some upgrades.
Like a X5670 (6 cores, 2.93 GHz, 6.40 GT/s, TDP 95 W), 50% more cores with 25% less TDP so the lower electricity bill would quick make up for the higher price. And depending on memory configuration a X5670 can take up to DDR3-1333 (modules like 32GB DDR2-1333, 30nm, 4.5W).
ID: 51827 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Andrew McFarland
Avatar

Send message
Joined: 21 Sep 11
Posts: 41
Credit: 185,458,736
RAC: 0
Message 51843 - Posted: 5 Dec 2011, 20:34:35 UTC - in response to Message 51805.  

NEW SERVER SPECIFICATIONS:

Memory: 64GB Memory (8x8GB), 800MHz 2R LV RDIMMs at Std Volt for 1 Proc, Optimized

Processor: Intel® Xeon® X5647 2.93GHz, 12M Cache, 5.86 GT/s QPI, 4C



Nice CPU choice. However, thats an strange number of DIMMs for a triple channel cpu, unless you've picked one of those entry level boards (like the TYAN S7002AG2NR) that has 8 slots (3+1 per CPU.) On the other hand, that board requires both CPU's to be popluated in order to utilize all 8 slots, and you only mention one here. Are you going dual cpu (not core) or single? For that matter, what motherboard are you using? Looking forward to the new arrival! :-)
ID: 51843 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Dragon7404

Send message
Joined: 9 Dec 09
Posts: 2
Credit: 290,169
RAC: 0
Message 51851 - Posted: 6 Dec 2011, 21:07:46 UTC

That outage during Thanksgiving really shot my credit average to china the hard way!!! It is back up but not any where near what it use to be.I was looking to be over 5k Credit average but that will not happen any time soon now.[/b]
ID: 51851 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
BarryAZ

Send message
Joined: 1 Sep 08
Posts: 520
Credit: 302,525,188
RAC: 0
Message 51852 - Posted: 6 Dec 2011, 22:38:23 UTC

Here's hoping that the new server does resolve the frequent outage scenario (such as the one which started about 5 or 6 hours ago and appears set to continue onto tomorrow).
ID: 51852 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile TimeRanger

Send message
Joined: 31 Oct 10
Posts: 83
Credit: 38,632,375
RAC: 0
Message 51859 - Posted: 8 Dec 2011, 20:32:38 UTC

I still wish there was a way for us to have more tasks "in the bank" to carry us through the outtages - whether they be system wide crashes or just routine maintenance.
ID: 51859 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile banditwolf
Avatar

Send message
Joined: 12 Nov 07
Posts: 2425
Credit: 524,164
RAC: 0
Message 51860 - Posted: 8 Dec 2011, 21:14:10 UTC - in response to Message 51859.  

I still wish there was a way for us to have more tasks "in the bank" to carry us through the outtages - whether they be system wide crashes or just routine maintenance.

The n-body have ran longer lately, but that doesn't help out gpu's at all.
Doesn't expecting the unexpected make the unexpected the expected?
If it makes sense, DON'T do it.
ID: 51860 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Jay Pound

Send message
Joined: 11 Oct 09
Posts: 1
Credit: 249,275,526
RAC: 204
Message 51895 - Posted: 10 Dec 2011, 4:42:27 UTC

Hey Guys,
Please don't use RAID 5, switch over to RAID 10. I build enterprise systems for a living, RAID 5 is dead dudes, espically on a database server, Why you break my heart?

I just built my first tiny supercomputer for m@h, quad gpu, socket 2011 six core, another one still to be built in the box(es), the milkyway project just cant feed the gpu's in the system as they take less than a minute to compute a single workunit per gpu core. I'm turning around 4 gpu workunits in 45-55secs, the server just cannot feed the system fast enough.

I'm hoping the findings from milkyway@home will eventually lead to a nbody simulation/equation that helps expose/prove the cyclic universe theory!!! My current Kraken is 13.6tflops strong, too bad it can currently only compute gpu workunits 5 mins out of every hour! Thanks for providing an awesome project, and some faith in universal science... I just cannot stomach all the tflops going to collatz!
-Jay Pound
ID: 51895 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Hendrik

Send message
Joined: 4 Dec 10
Posts: 1
Credit: 1,027,856
RAC: 0
Message 51921 - Posted: 11 Dec 2011, 14:09:26 UTC

I was thinking same, changeing to RAID 10 gives speed and redundancy. Raid 5 depending how many HDDs you use, latency goes up. Write speed (4 hhds) 2X and read speed 4x.
ID: 51921 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : News : New Server

©2024 Astroinformatics Group