New Data from AO

Message boards : News : New Data from AO
Message board moderation

To post messages, you must log in.

1 · 2 · 3 · 4 · Next

AuthorMessage
Jeff Cobb Project Donor
Volunteer moderator
Project administrator
Project developer
Project scientist

Send message
Joined: 1 Mar 99
Posts: 122
Credit: 40,367
RAC: 0
United States
Message 1609524 - Posted: 5 Dec 2014, 17:54:33 UTC
Last modified: 5 Dec 2014, 17:55:48 UTC

The four disks from AO have arrived - about 5TB of data. I have queued them up for pre-processing (radar blanking) so new SETI@Home workunits should be going out soon.
ID: 1609524 · Report as offensive
Profile Rene
Volunteer tester

Send message
Joined: 30 Jan 01
Posts: 25
Credit: 690,805
RAC: 0
Netherlands
Message 1609525 - Posted: 5 Dec 2014, 18:04:37 UTC

That´s great news, looking forward for some more SETI work!
ID: 1609525 · Report as offensive
Profile Mr. Kevvy Crowdfunding Project Donor*Special Project $250 donor
Volunteer moderator
Volunteer tester
Avatar

Send message
Joined: 15 May 99
Posts: 3806
Credit: 1,114,826,392
RAC: 3,319
Canada
Message 1609526 - Posted: 5 Dec 2014, 18:05:18 UTC
Last modified: 5 Dec 2014, 18:05:51 UTC

Nice... finally dry spell coming to an end (it will be difficult leaving E@H home though with the higher RAC!)

Two queries:

1) Is there enough storage (ie drive units) with this sudden amount of work? (Money burning hole in my pocket... heh.)

2) Is there going to be any Green Bank work mixed in, now or in the near future?
ID: 1609526 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11415
Credit: 29,581,041
RAC: 66
United States
Message 1609530 - Posted: 5 Dec 2014, 18:32:30 UTC

The Cricket graph is showing a big spike. That might be the new data, I hope. A question I have is how long will 5 TB of data last?
ID: 1609530 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1609532 - Posted: 5 Dec 2014, 18:42:45 UTC - in response to Message 1609530.  
Last modified: 5 Dec 2014, 18:43:23 UTC

5TB of data would theoretically be:

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 kilobytes
5497558138880 KB / 365 KB = 1,5061,803,120 Multibeam tasks (average 365 KB task size).

And if there were APs to split (which I doubt):
5497558138880 KB / (8 x 1024 x 1024) = 655,360 AP tasks (8 MB task size).

But there's probably some loss there. And perhaps some parts unusable.
ID: 1609532 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1609538 - Posted: 5 Dec 2014, 19:08:33 UTC - in response to Message 1609532.  
Last modified: 5 Dec 2014, 19:12:38 UTC

5TB of data would theoretically be:

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 kilobytes
5497558138880 KB / 365 KB = 1,5061,803,120 Multibeam tasks (average 365 KB task size).
...

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 BYTES

Multibeam tasks actually only have 256K (262144 Bytes) data, but it's encoded in a fashion similar to uue which makes it larger. Also, splitting uses about a 20% overlap between WU groups so a better estimate would be

(5497558138880 / 262144) * 1.2 = 25,165,824 Multibeam workunits.

It usually takes about 2.1 tasks on average per Multibeam workunit, so 52,848,230 tasks or so is my estimate.
                                                                  Joe
ID: 1609538 · Report as offensive
Anthony Ayiomamitis

Send message
Joined: 2 May 04
Posts: 19
Credit: 739,837
RAC: 0
Greece
Message 1609541 - Posted: 5 Dec 2014, 19:24:41 UTC - in response to Message 1609526.  

Nice... finally dry spell coming to an end (it will be difficult leaving E@H home though with the higher RAC!)

A very pleasant surprise which I noticed as well. Is there some online link which provides the RAC on a per project basis for comparison purposes?
ID: 1609541 · Report as offensive
Profile Francis Noel
Avatar

Send message
Joined: 30 Aug 05
Posts: 452
Credit: 142,832,523
RAC: 94
Canada
Message 1609542 - Posted: 5 Dec 2014, 19:25:29 UTC - in response to Message 1609538.  

Joy !
mambo
ID: 1609542 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14679
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1609544 - Posted: 5 Dec 2014, 19:28:44 UTC - in response to Message 1609538.  

5TB of data would theoretically be:

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 kilobytes
5497558138880 KB / 365 KB = 1,5061,803,120 Multibeam tasks (average 365 KB task size).
...

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 BYTES

Multibeam tasks actually only have 256K (262144 Bytes) data, but it's encoded in a fashion similar to uue which makes it larger. Also, splitting uses about a 20% overlap between WU groups so a better estimate would be

(5497558138880 / 262144) * 1.2 = 25,165,824 Multibeam workunits.

It usually takes about 2.1 tasks on average per Multibeam workunit, so 52,848,230 tasks or so is my estimate.
                                                                  Joe

Let's say that the loose change (2.8 million MB tasks) vanishes into the 'out in the field' caches - so to an order of magnitude, 50 million tasks and we process 50 thousand an hour.

So, around a thousand hours of work - can't estimate closer than that without any knowledge of the ARs of the recordings. 6 weeks of work should keep us warm through the holidays.
ID: 1609544 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11415
Credit: 29,581,041
RAC: 66
United States
Message 1609545 - Posted: 5 Dec 2014, 19:32:18 UTC - in response to Message 1609544.  

6 weeks of work should keep us warm through the holidays

Ah, an answer I can understand. Thanx
ID: 1609545 · Report as offensive
Profile JaundicedEye
Avatar

Send message
Joined: 14 Mar 12
Posts: 5375
Credit: 30,870,693
RAC: 1
United States
Message 1609559 - Posted: 5 Dec 2014, 20:28:57 UTC - in response to Message 1609534.  

That's all good and well Jeff, but how about the rebuilding of the AP database? We haven't heard a word about that since Dec 2, when Eric predicted that it could be finished on Saturday, and that he should post updates during the week (which didn't happen).


Saturday is tomorrow.........I'm not going to look gift horses or WU's in the mouth. Anything to crunch is a plus.

"Sour Grapes make a bitter Whine." <(0)>
ID: 1609559 · Report as offensive
Alex2
Volunteer tester

Send message
Joined: 2 Nov 12
Posts: 24
Credit: 11,798,226
RAC: 0
Canada
Message 1609565 - Posted: 5 Dec 2014, 20:47:41 UTC

6 weeks of work

It looks like it is already finished.
No more are available.
ID: 1609565 · Report as offensive
Profile Wiggo
Avatar

Send message
Joined: 24 Jan 00
Posts: 36764
Credit: 261,360,520
RAC: 489
Australia
Message 1609570 - Posted: 5 Dec 2014, 20:56:42 UTC - in response to Message 1609565.  

6 weeks of work

It looks like it is already finished.
No more are available.

Now how do you work that out when there are 3 files currently being split as we speak? :-O

Cheers.
ID: 1609570 · Report as offensive
Profile Andy Lee Robinson
Avatar

Send message
Joined: 8 Dec 05
Posts: 630
Credit: 59,973,836
RAC: 0
Hungary
Message 1609571 - Posted: 5 Dec 2014, 21:00:28 UTC
Last modified: 5 Dec 2014, 21:00:55 UTC

Yay! At last the wheels are moving again!
http://setistats.haveland.com/
ID: 1609571 · Report as offensive
Profile Jord
Volunteer tester
Avatar

Send message
Joined: 9 Jun 99
Posts: 15184
Credit: 4,362,181
RAC: 3
Netherlands
Message 1609572 - Posted: 5 Dec 2014, 21:04:53 UTC - in response to Message 1609538.  
Last modified: 5 Dec 2014, 21:05:34 UTC

5 x 1024 x 1024 x 1024 x 1024 = 5,497,558,138,880 BYTES

lol, did I go one too many? Ah yes. ;-)
So divide by 1024...
ID: 1609572 · Report as offensive
Profile Temporelo
Avatar

Send message
Joined: 18 May 99
Posts: 90
Credit: 31,082,860
RAC: 22
Canada
Message 1609584 - Posted: 5 Dec 2014, 21:32:09 UTC

Wow, great news !
Even though I run another project, my preference and loyalty has always been for Seti.
Thank you for the team of Seti for their efforts.

On ne peut voir au-delà des choix qu'on ne comprend pas.
ID: 1609584 · Report as offensive
Profile S@NL Etienne Dokkum
Volunteer tester
Avatar

Send message
Joined: 11 Jun 99
Posts: 212
Credit: 43,822,095
RAC: 0
Netherlands
Message 1609592 - Posted: 5 Dec 2014, 21:46:48 UTC

Thanks, 200 MB tasks in progress as we speak here :D

that's a great Sinterklaas(Dutch traditional holiday where gifts are given to children by a Santa Claus kind of folk hero) present !
ID: 1609592 · Report as offensive
Profile Braingasim
Avatar

Send message
Joined: 31 Aug 99
Posts: 17
Credit: 2,580,690
RAC: 6
Netherlands
Message 1609603 - Posted: 5 Dec 2014, 22:18:50 UTC

Woohoo, more work to process!
ID: 1609603 · Report as offensive
Profile cwburch

Send message
Joined: 23 Jul 00
Posts: 67
Credit: 61,429,161
RAC: 27
United States
Message 1609626 - Posted: 5 Dec 2014, 23:56:01 UTC

Well, I see on the server status page there's no backlog of WU's to send, no matter how often I update the page. All of my computers except one have SETI WUs now. Of course, the one that doesn't is my best one with 2 6-core NUMA processors. So I've got 24 virtual processors sitting idle if it wasn't for the Bitcoin fundraiser. I just manually grab a load of bitcoin, process it, check SETI and if no tasks there, grab another load of BTC WUs. Hopefully, I'll get lucky before bedtime and grab enough SETI WUs to work through the night.
My goal for the year is 20M credits, and I'm soooooooooo close. So I want this particular machine working SETI. I figure once I reach that 20M goal, then I'll let all of my computer work the BTC WUs until the end of the campaign.
ID: 1609626 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1609635 - Posted: 6 Dec 2014, 1:12:58 UTC - in response to Message 1609626.  

Well, I see on the server status page there's no backlog of WU's to send, no matter how often I update the page. All of my computers except one have SETI WUs now. Of course, the one that doesn't is my best one with 2 6-core NUMA processors. So I've got 24 virtual processors sitting idle if it wasn't for the Bitcoin fundraiser. I just manually grab a load of bitcoin, process it, check SETI and if no tasks there, grab another load of BTC WUs. Hopefully, I'll get lucky before bedtime and grab enough SETI WUs to work through the night.
My goal for the year is 20M credits, and I'm soooooooooo close. So I want this particular machine working SETI. I figure once I reach that 20M goal, then I'll let all of my computer work the BTC WUs until the end of the campaign.

The server status page only gets generated once every 10 or 20 minutes. No matter how many times you refresh it, take note of the important bit at the top: "[As of 6 Dec 2014, 1:00:10 UTC]." The next refresh will either be 1:10, or 1:20. It is supposed to be a 10-minute interval, but sometimes it takes more than 10 minutes for the script to run all the way through generating all the values, so it just waits until the next 10 interval.

Regarding having trouble getting work.. Some people have found that having a huge cache setting, or requesting a very large number of seconds of work, the scheduler tends to not want to give you WUs. Change your cache settings down to something small, like 0.1 days. Once you get a few WUs, then you can go back to large settings and it seems to be a little more cooperative.

That aside.. you also have to remember there are 100,000+ hungry machines DDoSing the server asking for work. The feeder queue (what's available when the scheduler assigns work to somebody) can only hold 200 tasks at a time, and only refills every 2-5 seconds, so it is very much luck of the draw as to when your request gets processed versus what's available in the feeder at that same moment.

Just be patient.. it'll be a few hours before things start to settle down.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1609635 · Report as offensive
1 · 2 · 3 · 4 · Next

Message boards : News : New Data from AO


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.