Advanced search

Message boards : News : Major server upgrade!

Author Message
Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 45290 - Posted: 17 Nov 2016 | 10:48:09 UTC

Dear cruchers!

the GPUGRID server you know and (I guess?) love has done a huge service to science. Thanks to your support, the experiments we have being running grew in size and scope... AND we have enough ideas for the future to keep this trend.

So, we decided to significantly upgrade the capabilities of our server - and a new is almost in place. Likely next Monday, we'll stop the server for some hours, then perform the transition. We now need your patience, because the switch-over will require some down time (and hiccups, possibly).

But we are definitely looking forwards for the new expanded experimental capacity! We'll keep you informed!


Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45291 - Posted: 17 Nov 2016 | 11:21:33 UTC - in response to Message 45290.

Thank you Toni,

Sounds like good news and since you think its worth the investment it makes me more confident that the work and my contribution to it is worthwhile.

Downtime and teething problems are to be expected in this case.

kain
Send message
Joined: 3 Sep 14
Posts: 152
Credit: 641,182,245
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 45292 - Posted: 17 Nov 2016 | 11:38:19 UTC

Could you give us the old and the news specs?

John C MacAlister
Send message
Joined: 17 Feb 13
Posts: 181
Credit: 144,871,276
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 45293 - Posted: 17 Nov 2016 | 12:13:51 UTC

Thanks, Toni.

Bring on the new beast....
____________
John

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 184
Credit: 222,376,233
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 45294 - Posted: 17 Nov 2016 | 13:51:49 UTC
Last modified: 17 Nov 2016 | 13:54:33 UTC

Good news. My hope is that somehow the ever annoying download issue will be gone when the new server is fully operational.

WPrion
Send message
Joined: 30 Apr 13
Posts: 87
Credit: 1,065,409,111
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45295 - Posted: 17 Nov 2016 | 14:12:38 UTC - in response to Message 45294.
Last modified: 17 Nov 2016 | 14:13:52 UTC

"My hope is that somehow the ever annoying download issue will be gone when the new server is fully operational."

+100

PappaLitto
Send message
Joined: 21 Mar 16
Posts: 511
Credit: 4,617,042,755
RAC: 0
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 45298 - Posted: 17 Nov 2016 | 14:56:42 UTC - in response to Message 45295.
Last modified: 17 Nov 2016 | 14:57:37 UTC

"My hope is that somehow the ever annoying download issue will be gone when the new server is fully operational."

+100


+1000000

Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 45301 - Posted: 17 Nov 2016 | 15:55:34 UTC - in response to Message 45298.

Yes, our commitment is steady. The computing power you are donating provides a scientific tool whose importance can not be overstated.

The most important upgrade is the disks - storage space was somewhat limiting, in the sense that we had to pay attention to not having too many experiments running at the same time. (And note that the server only works as a temporary buffer!)

[VENETO] boboviz
Send message
Joined: 10 Sep 10
Posts: 158
Credit: 388,132
RAC: 0
Level

Scientific publications
wat
Message 45302 - Posted: 17 Nov 2016 | 17:22:46 UTC

It's "only" a hw upgrade or there is also some sw updates (for example boinc server libraries)??

Jozef J
Send message
Joined: 7 Jun 12
Posts: 112
Credit: 1,118,845,172
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 45305 - Posted: 17 Nov 2016 | 19:35:21 UTC - in response to Message 45298.

"My hope is that somehow the ever annoying download issue will be gone when the new server is fully operational."

+100


+1000000


+10*291 :-)

Profile Maurice Goulois
Avatar
Send message
Joined: 22 Feb 09
Posts: 10
Credit: 103,904,673
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45309 - Posted: 18 Nov 2016 | 2:21:46 UTC
Last modified: 18 Nov 2016 | 2:22:59 UTC

Very good news, hoping it goes well :)

Mark West
Send message
Joined: 7 Oct 12
Posts: 5
Credit: 416,078,317
RAC: 0
Level
Gln
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45310 - Posted: 18 Nov 2016 | 2:54:05 UTC

Good news for sure! I look forward to crunching even more work units. I might even bring a second GPU on line to celebrate.

Patrick Massey
Send message
Joined: 27 Dec 11
Posts: 1
Credit: 42,164,562
RAC: 0
Level
Val
Scientific publications
watwatwatwatwatwatwatwatwatwat
Message 45311 - Posted: 18 Nov 2016 | 5:29:30 UTC

Congrats guys! This is good news for both us (the users) and your team[s] as well!

And yes, specs would be most interesting to see (they don't want to tell us it's actually being hosted on an old HP-UX box with a PA-RISC CPU, or maybe an IBM AS/400...).
____________
"Apollo was astonished, Dionysus thought me mad..."

Profile Michael H.W. Weber
Send message
Joined: 9 Feb 16
Posts: 71
Credit: 607,916,391
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwat
Message 45313 - Posted: 18 Nov 2016 | 9:17:35 UTC

To my point of view you do not have a major problem with the storage devices but rather with bandwidth. You can't imagine how slow data transfer from your server is compared to other DC projects and how often it just breaks down and causes a reduction in project productivity.

Michael.
____________
President of Rechenkraft.net - Germany's first and largest distributed computing organization.

Profile Francois Normandin
Send message
Joined: 8 Mar 11
Posts: 71
Credit: 654,432,613
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 45314 - Posted: 18 Nov 2016 | 10:40:42 UTC
Last modified: 18 Nov 2016 | 10:43:59 UTC

Congrats! :)

Profile Francois Normandin
Send message
Joined: 8 Mar 11
Posts: 71
Credit: 654,432,613
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 45315 - Posted: 18 Nov 2016 | 10:44:16 UTC - in response to Message 45305.

+googolplex :)

Profile Francois Normandin
Send message
Joined: 8 Mar 11
Posts: 71
Credit: 654,432,613
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwat
Message 45316 - Posted: 18 Nov 2016 | 11:04:20 UTC - in response to Message 45305.

"My hope is that somehow the ever annoying download issue will be gone when the new server is fully operational."

+100


+1000000


+10*291 :-)


+googolplex :D

PappaLitto
Send message
Joined: 21 Mar 16
Posts: 511
Credit: 4,617,042,755
RAC: 0
Level
Arg
Scientific publications
watwatwatwatwatwatwatwat
Message 45317 - Posted: 18 Nov 2016 | 12:23:31 UTC

What is your internet connnection speed? Because ultimately no matter how fast your internal network is, the bandwidth won't increase unless you have faster internet.

Profile Michael H.W. Weber
Send message
Joined: 9 Feb 16
Posts: 71
Credit: 607,916,391
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwat
Message 45318 - Posted: 18 Nov 2016 | 13:56:28 UTC - in response to Message 45317.

What is your internet connnection speed? Because ultimately no matter how fast your internal network is, the bandwidth won't increase unless you have faster internet.

The problem is at the GPUGRID server's end - believe me...

:)

Michael.
____________
President of Rechenkraft.net - Germany's first and largest distributed computing organization.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2343
Credit: 16,201,255,749
RAC: 7,520
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45320 - Posted: 18 Nov 2016 | 16:02:31 UTC - in response to Message 45319.

The virus spreading bots are also happy about the server upgrade.

Old man
Send message
Joined: 24 Jan 09
Posts: 42
Credit: 16,676,387
RAC: 0
Level
Pro
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45321 - Posted: 18 Nov 2016 | 16:04:25 UTC - in response to Message 45313.

To my point of view you do not have a major problem with the storage devices but rather with bandwidth. You can't imagine how slow data transfer from your server is compared to other DC projects and how often it just breaks down and causes a reduction in project productivity.

Michael.


Try Seti@home project and you find very slow internet connection...and slow downloads.

[CSF] Thomas H.V. DUPONT
Send message
Joined: 20 Jul 14
Posts: 732
Credit: 100,630,366
RAC: 0
Level
Cys
Scientific publications
watwatwatwatwatwatwatwat
Message 45346 - Posted: 21 Nov 2016 | 8:15:52 UTC

Good news, Toni!
Thanks for the heads-up :)
____________
[CSF] Thomas H.V. Dupont
Founder of the team CRUNCHERS SANS FRONTIERES 2.0
www.crunchersansfrontieres

Profile Michael H.W. Weber
Send message
Joined: 9 Feb 16
Posts: 71
Credit: 607,916,391
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwat
Message 45380 - Posted: 24 Nov 2016 | 15:34:56 UTC - in response to Message 45321.
Last modified: 24 Nov 2016 | 15:39:05 UTC

Try Seti@home project and you find very slow internet connection...and slow downloads.

And because SETI's network performance is bad (which I actually can't confirm at all), GPUGRID has to be, too?

The problem is that once a GPUGRID download is interrupted (which happens on a daily basis), no further work is delivered for a day. And that, in turn, reduces the project's overall throughput.

As a consequence, at present, at our end over here, it is advisable to have a second GPU project running in parallel to GPUGRID just because GPUGRID's reliability in terms of work delivery is extremely insufficient. This other project then of course competes for the GPU cycles...

Michael.
____________
President of Rechenkraft.net - Germany's first and largest distributed computing organization.

Betting Slip
Send message
Joined: 5 Jan 09
Posts: 670
Credit: 2,498,095,550
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45415 - Posted: 28 Nov 2016 | 1:13:18 UTC - in response to Message 45290.
Last modified: 28 Nov 2016 | 1:15:54 UTC

Dear cruchers!

the GPUGRID server you know and (I guess?) love has done a huge service to science. Thanks to your support, the experiments we have being running grew in size and scope... AND we have enough ideas for the future to keep this trend.

So, we decided to significantly upgrade the capabilities of our server - and a new is almost in place. Likely next Monday, we'll stop the server for some hours, then perform the transition. We now need your patience, because the switch-over will require some down time (and hiccups, possibly).

But we are definitely looking forwards for the new expanded experimental capacity! We'll keep you informed!




Please give us a more firm ETA for server upgrade. I can't help but think you have let us down again both in server upgrade and keeping us informed about said upgrade and project as a whole.

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 1576
Credit: 5,598,486,851
RAC: 8,769,820
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 45417 - Posted: 28 Nov 2016 | 9:11:17 UTC - in response to Message 45290.

Likely next Monday ...

Does the empty ready-to-send queue imply that you meant *this* Monday?

Toni
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist
Send message
Joined: 9 Dec 08
Posts: 1006
Credit: 5,068,599
RAC: 0
Level
Ser
Scientific publications
watwatwatwat
Message 45430 - Posted: 29 Nov 2016 | 11:37:38 UTC - in response to Message 45417.

Preparations for the move are ongoing. Another reason why we can't fill up the queue right now is that the old disk is at it capacity limit.

By the way, old specs:

Intel(R) Xeon(R) CPU X3470 @ 2.93GHz, 4 cores
24 GB of RAM
~2 TB of storage
Redundant storage, power, etc.

New specs:

Intel(R) Xeon(R) CPU E5-2620 v3 @ 2.40GHz, 6 cores x 2 cpus
32 GB of RAM
~6 TB of storage
Also redundant.

Both are Dell Poweredge. The OS will be updated as well, implying that the recompiled BOINC server may require some tweak period.



kain
Send message
Joined: 3 Sep 14
Posts: 152
Credit: 641,182,245
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwat
Message 45433 - Posted: 29 Nov 2016 | 12:27:13 UTC

Huge upgrade! Great :D

Profile Michael H.W. Weber
Send message
Joined: 9 Feb 16
Posts: 71
Credit: 607,916,391
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwat
Message 45563 - Posted: 3 Dec 2016 | 11:11:03 UTC

Upload took again hours to complete with only a few bytes of result data compared to other projects.
And: Since days, WUs are not delivered again.
Very disappointing.

Michael.
____________
President of Rechenkraft.net - Germany's first and largest distributed computing organization.

Erich56
Send message
Joined: 1 Jan 15
Posts: 1090
Credit: 6,603,906,926
RAC: 21,893,126
Level
Tyr
Scientific publications
watwatwatwatwatwatwatwatwat
Message 45577 - Posted: 3 Dec 2016 | 19:27:13 UTC

Toni, any rough estimation as to when new WUs will be available for download ?

Post to thread

Message boards : News : Major server upgrade!

//