Holiday Checkup (Dec 23 2007)

Message boards : Technical News : Holiday Checkup (Dec 23 2007)

Author	Message
Matt Lebofsky Volunteer moderator Project administrator Project developer Project scientist Send message Joined: 1 Mar 99 Posts: 1383 Credit: 74,079 RAC: 0	Message 693979 - Posted: 23 Dec 2007, 19:05:25 UTC
	Quick note: We never really did recover from the science database issues from a couple days ago due to DOS'ing ourselves with fast workunits. Whatever. We chose to let things naturally pass through the system. Kinda like kidney stones. Meanwhile, one of the assimilators is failing with a brand new error. If any of us have time we'll try to check into that over the coming days, but we may be out of luck until we're all in the lab doing "extreme debugging" together on Wednesday. Hang in there! - Matt ____________ -- BOINC/SETI@home network/web/science/development person -- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude
	ID: 693979 ·

Dr. C.E.T.I. Send message Joined: 29 Feb 00 Posts: 15972 Credit: 679,129 RAC: 105	Message 693995 - Posted: 23 Dec 2007, 21:03:36 UTC
	. . . been waiting / crunchin' for almost 8 Years - another day passes [c'est la vie] ;) - Thanks for the Post Matt and Happy Holidays Sir . . . richard ____________ BOINC Wiki . . . Science Status Page . . .
	ID: 693995 ·

Richard Haselgrove Volunteer tester Send message Joined: 4 Jul 99 Posts: 8161 Credit: 44,448,641 RAC: 11,551	Message 693998 - Posted: 23 Dec 2007, 21:57:59 UTC
	Just done a quick check on all WUs logged here. 530 out of 631 (83.99%) of all tasks issued after 12:00:00 UTC 22 Dec 2007 are shown as Angle Range 2.46 to 2.64 That compares with 399 out of 9766 (4.09%) of tasks issued in the previous four months. Either something very odd about that data, or something very odd has happened to the splitters. I think it's a very sensible decision to let it runs its course over the weekend, and I don't begrudge you your holiday break - Happy Christmas, and have a great time as far away from the lab as possible! But when you come in for that "extreme debugging" session (as early as Wednesday? Shame on you!), please have a look for causes as well as coping strategies.
	ID: 693998 ·

lwblack1051 Send message Joined: 16 Dec 07 Posts: 1 Credit: 40,837 RAC: 0	Message 694106 - Posted: 24 Dec 2007, 6:09:54 UTC - in response to Message 694078.
	Just done a quick check on all WUs logged here. 530 out of 631 (83.99%) of all tasks issued after 12:00:00 UTC 22 Dec 2007 are shown as Angle Range 2.46 to 2.64 That compares with 399 out of 9766 (4.09%) of tasks issued in the previous four months. Either something very odd about that data, or something very odd has happened to the splitters. I think it's a very sensible decision to let it runs its course over the weekend, and I don't begrudge you your holiday break - Happy Christmas, and have a great time as far away from the lab as possible! But when you come in for that "extreme debugging" session (as early as Wednesday? Shame on you!), please have a look for causes as well as coping strategies. I have to wonder what is going on as well..... The new 02dc06aa series also appears to be coming up with all shorties....something doesn't seem right. Allot of project server down messages of late. Spending more and more processor time on getting can't connect messages.
	ID: 694106 ·

Jan Schotsmans Send message Joined: 27 Oct 00 Posts: 98 Credit: 92,693 RAC: 0	Message 694134 - Posted: 24 Dec 2007, 7:52:57 UTC - in response to Message 694111.
	Allot of project server down messages of late. Spending more and more processor time on getting can't connect messages. 'S 'cuz the splitters cannot create enough WUs to keep up with the demand from the host computers..... They are being crunched and reported as fast as they can be created.... And the kitties say....'Stay tuned for further developements.......' Lets hope the next tape is nearly all 80+ point WU's so that the splitters can create a new WU queue. These 17 pointer ones are finishing in just over 20 minutes for me, that makes 2 units per 20 minutes on my C2D laptop, SETI would need to have 144 WU's ready per C2D level machine to keep this up, with a few million hosts running the app its obvious why the splitters can't keep up.
	ID: 694134 ·

Rev. Tim Olivera Send message Joined: 15 Jan 06 Posts: 20 Credit: 1,717,714 RAC: 0	Message 694158 - Posted: 24 Dec 2007, 10:50:44 UTC - in response to Message 693979.
	True to from just like on every long holiday no work! But the thing that made me really laugh this time was a letter asking for more folks to help with your program!! You can't keep the folks you hve now in work so you need more folks to leave there computers on to do nothing but use power and wear out parts!!! For 3 days now BOINC is Idle!! Keep up the good work guys don't let me down for Easter ot the 3 day in Feb I love to have my computers sucking up power and wearing out parts for nothing!!! rev. Tim Olivera ____________
	ID: 694158 ·

RandyC Send message Joined: 20 Oct 99 Posts: 714 Credit: 1,704,345 RAC: 0	Message 694164 - Posted: 24 Dec 2007, 11:28:34 UTC - in response to Message 694158.
	True to from just like on every long holiday no work! But the thing that made me really laugh this time was a letter asking for more folks to help with your program!! You can't keep the folks you hve now in work so you need more folks to leave there computers on to do nothing but use power and wear out parts!!! For 3 days now BOINC is Idle!! Keep up the good work guys don't let me down for Easter ot the 3 day in Feb I love to have my computers sucking up power and wearing out parts for nothing!!! rev. Tim Olivera If you want to keep your systems busy when SAH is out of work, then attach to a backup project. S@H has specifically stated they WILL NOT always have work available and advises users to attach to a backup project for this exact situation.
	ID: 694164 ·

Richard Haselgrove Volunteer tester Send message Joined: 4 Jul 99 Posts: 8161 Credit: 44,448,641 RAC: 11,551	Message 694165 - Posted: 24 Dec 2007, 11:34:39 UTC - in response to Message 694164.
	True to from just like on every long holiday no work! But the thing that made me really laugh this time was a letter asking for more folks to help with your program!! You can't keep the folks you hve now in work so you need more folks to leave there computers on to do nothing but use power and wear out parts!!! For 3 days now BOINC is Idle!! Keep up the good work guys don't let me down for Easter ot the 3 day in Feb I love to have my computers sucking up power and wearing out parts for nothing!!! rev. Tim Olivera If you want to keep your systems busy when SAH is out of work, then attach to a backup project. S@H has specifically stated they WILL NOT always have work available and advises users to attach to a backup project for this exact situation. For once - just once - I think the Rev. Tim has a point. The "please help our project" letter doesn't say "please help our backup project". Co-ordination and timing, that's the thing: if you ask for help, it's only polite to be ready to accept it.
	ID: 694165 ·

Jan Schotsmans Send message Joined: 27 Oct 00 Posts: 98 Credit: 92,693 RAC: 0	Message 694179 - Posted: 24 Dec 2007, 13:04:21 UTC
	Instead of focussing and whining about the CPU cycle request letter sent arround, focus on the "WE NEED FUNDS" one that went arround recently too. These guys are working severely underpaid for their skillevel, with maybe a quarter of the hardware, half the bandwidth and barely any of the cash they need to keep operating.
	ID: 694179 ·

Neil Blaikie Volunteer tester Send message Joined: 17 May 99 Posts: 142 Credit: 6,422,726 RAC: 1,300	Message 694191 - Posted: 24 Dec 2007, 13:45:17 UTC
	Am sure the guys get a ton of messages for help every day and physically don't have time to check every single one. If they did then they would have too much time on their hands. I have just started to downloaded new tasks, I have a fast computer so it turns a task around in a little near 2 hours. I don't expect them to download quick, currently getting 2.26kpbs with a majority of the tasks in line to download pending. I don't have a backup project so will just have to sit it out and get work when I can, am sure the guys at Berkeley would appreciate a few days off to relax at home and not have to worry about the work line being empty. No work, ah well, will get it when it comes. Happy Holidays all! ____________
	ID: 694191 ·

John Neale Volunteer tester Send message Joined: 16 Mar 00 Posts: 478 Credit: 2,638,561 RAC: 1,986	Message 694201 - Posted: 24 Dec 2007, 14:31:59 UTC - in response to Message 694158. Last modified: 24 Dec 2007, 14:33:24 UTC
	True to from (sic) just like on every long holiday no work! Not true, Tim. The SETI servers have never churned out as many new workunits as they have in the past few days. They're actually at their limit, demonstrating what they're capable of. For 3 days now BOINC is Idle!! Not true, Tim. Your two active rigs have been receiving and returning work for the past three days. Rig 1 Rig 2 I've taken issue with you before for your incessant whingeing about this project. It's not justified, in my opinion, but using untruths to propagate your message is disingenuous. ____________
	ID: 694201 ·

Neil Blaikie Volunteer tester Send message Joined: 17 May 99 Posts: 142 Credit: 6,422,726 RAC: 1,300	Message 694239 - Posted: 24 Dec 2007, 15:58:33 UTC
	Maybe if more people take heed of the donation drive that the guys at Berkeley spent time sending to us all and more people sent money then perhaps things would be a bit better. There will always be the "few" that have things to say regarding the running and subsequent problems associated with a massive project like this. Come on though people. It's A) Christmas and the guys at Berkeley have to work / check in on things, for what?, to keep us "joe public" happy. As someone mentioned earlier, attach to a backup project if you want to keep crunching workunits. B) Anyone who runs servers on a large scale will know that sometimes things don't run smoothly and you get hiccups. Matt and everyone at Berkeley does a fantastic job to keep everything running as smoothly as possible and they are human like the rest of us. They are allowed to have a break occasionally. With the resources they have and behind the scenes stuff that goes on that we don't see / get to hear about. They should be given as much credit as they can. No donations = ultimately no project and hundreds of hours of valuable time wasted due to lack of funding. Enjoy the holidays and get a well deserved break. Thanks for all the hard work as always. ____________
	ID: 694239 ·

Chris S Volunteer tester Send message Joined: 19 Nov 00 Posts: 28945 Credit: 7,811,999 RAC: 15,675	Message 694264 - Posted: 24 Dec 2007, 17:20:50 UTC
	@Matt I, and I am sure many others, would like to wish you the compliments of the season, and hope that you have a very happy christmas. I may have serious issues with others, this does not apply to you. ____________ Damsel Rescuer, Kitty Patron, Uli Fan, Julie Supporter, CAMRA ES99 Admirer, Raccoon Friend, IFAW, PETA, 5% Badge
	ID: 694264 ·

seti@elrcastor.com Volunteer tester Send message Joined: 30 Jan 00 Posts: 35 Credit: 4,847,711 RAC: 505	Message 694265 - Posted: 24 Dec 2007, 17:22:14 UTC Last modified: 24 Dec 2007, 17:22:41 UTC
	@Seti People Great Job on makeing things run so well, and have a great Christmas and New Years. ____________
	ID: 694265 ·

KWSN THE Holy Hand Grenade! Volunteer tester Send message Joined: 20 Dec 05 Posts: 1804 Credit: 6,693,945 RAC: 13,118	Message 694268 - Posted: 24 Dec 2007, 17:34:33 UTC Last modified: 24 Dec 2007, 17:35:28 UTC
	On the "clients recently connected" front, I'm getting "last updated: December 31 1969 16:00:00" at 0930 PST 12/24/07, so there are still date issues with the apache logs... ____________ .
	ID: 694268 ·

PhonAcq Send message Joined: 14 Apr 01 Posts: 1622 Credit: 21,478,187 RAC: 2,191	Message 694276 - Posted: 24 Dec 2007, 18:12:19 UTC - in response to Message 694236.
	And might you create a process whereby the splitters monitor things and automatically suspend a dataset that consists of such short work and return to it once they have built the cache up again? Not to say that it should not be processed, but could it be done in smaller doses so as not to decimate the ready to send cache and send the bandwidth to the limits? why couldn't the download process give the short time wu's to the slower clients and the longer (normal?) ones to the faster clients. I think biasing the work distribution could be used to even out the network loading, right?
	ID: 694276 ·

1 · 2 · Next

Message boards : Technical News : Holiday Checkup (Dec 23 2007)