1) Message boards : Number crunching : Is there a way ro remove hands from BOINC? (Message 28249) Posted 20 Sep 2004 by Christopher Hauber Post: You should note that just because it says "Ready to report" in the work tab, doesn't mean it hasn't already sent. Completed workunits are generally sent immediately (if your settings allow). As long as network access is not disabled, completed workunits will appear in the transfer tab until they are actually sent. Once they are sent, they stay in the work tab saying "ready to report" until you update the project. I don't know why that is, but just the way of things. You can confirm it by looking through the log and finding where things have been uploaded. > Management tasks multiply when you are running multiple computers. I like to > simplify these tasks. Therefore I have two questions: > > 1. Is there a way to make the "update" automatic so "ready to report" WUs go > by themselves? > > 2. Is there a way (a third-party-ad-on, perhaps) to see from a central > location the status and progress on each BOINC computer in the farm? With > SETI classic I use MSETIMON and SETIQUEUE - they give me a nice view into the > current state of each computer/WU and a nice way to look at historical and > statistical data for each machine? > > Thoughts? > > -mm- > >

2) Message boards : Number crunching : How Are Benchmarks Calculated? (Message 24939) Posted 11 Sep 2004 by Christopher Hauber Post: Hmm.. I knew the Celeron was the "value" line but I didn't realize just how inferior to the Pentiums they really were. My P3 1.0 GHz with 512 MB RAM only takes about an hour longer than your 1.0 GHz Celeron. I'm surprised there isn't a bigger difference there. As an added comparison my "2800+" (2400+ overclocked) processes work in about 2:50 on the low end of the 2800+ range (I say "range" because I have lowered the speed a couple times recently without it changing the 2800+ designation). Chris > I have a 2.0 Ghz Celeron with 256meg of Ram...and the benchmarks say a > workunit should take 6.5 hours and most units take 6.5 to 6.75 hours. > > Don > >

3) Message boards : Number crunching : Is anyone getting credit? (Message 24924) Posted 11 Sep 2004 by Christopher Hauber Post: Wow. If I did my math right that's roughly 2 days worth of backlog (when considering the incoming rate). That's a lot of work :) > There are roughly 97399 workunits pending validation. > > We are chewing through them at about 4000 an hour with an incoming rate of > about 2000 an hour. > > So its going to be awhile before we get through them all. >

4) Message boards : Number crunching : Merging PC´s (Message 24732) Posted 10 Sep 2004 by Christopher Hauber Post: I think it should work fine being merged. If you want to "unmerge" them, you should be able to do that by detaching/reattaching (make sure you have completed AND submitted all of your workunits before detaching!). You can probably also close BOINC, delete the computer ID number in the appropriate XML file, and then restart the program, forcing it to assign a new ID. I don't have BOINC installed at work anymore so I can't look this up unfortunately. Someone else can probably help in that regard. Chris > My guess is that it would get a new id number and life would go on. >

5) Message boards : Number crunching : Another new nice feature: edit forum preferences (Message 24693) Posted 10 Sep 2004 by Christopher Hauber Post: They are also waiting for some corrections to get through alpha testing -- like the validator. > As Rom said in his post, the new forum features were developed by someone not > on the Berkeley staff. I am sure Rom and the other devs are working away on > the other problems. >

6) Message boards : Number crunching : Another new nice feature: edit forum preferences (Message 24674) Posted 10 Sep 2004 by Christopher Hauber Post: Pretty cool. Appears that the photo I just turned into an avatar takes up exactly 4k (on disk - actual size is slightly smaller) when I reduced it down to 100x100 px. Hopefully I can find a photo of mine that reduces a little better though. The size (dimension) restrictions really don't give you much to work with. > > Ok thats a great feature, > > Now how do I get it on my page. I am full of questions when something > new is > > presented. I thank all that are willing to help. > > Just find a small jpg / png file (smaller than 4k in size) and upload it > through your account settings page on the site here. > ---

7) Message boards : Number crunching : Good news? (Message 23965) Posted 9 Sep 2004 by Christopher Hauber Post: Well, it's a little soon to say for sure, but from the latest news on the main page, it appears that we may be nearing the end of the catastrophic server problems. The snap server is still running stable. The found and fixed an infinite loop problem, which I think could potentially cause other system problems other than the validator simply not working as it should. While from what I understand, the SnapApp is is running with minimal "services" which I assume will later be turned on to find the problem, at least it is working properly as is, making it much easier to find out what kept causing it to hang. I know it's a bit early to tell on all this, but I get the feeling things are looking up at least a little. Hopefully that feeling is right.

8) Message boards : Number crunching : Work/Home/School preference don't work (Message 23929) Posted 9 Sep 2004 by Christopher Hauber Post: Check out these threads: http://setiweb.ssl.berkeley.edu/forum_thread.php?id=3354 http://setiweb.ssl.berkeley.edu/forum_thread.php?id=3888 > I have set different values for the number of WU's Boinc should download per > type (home/work/school). > Also, I set my computers to use one of these types, since they don't all need > the same number of WU's in the queue. > > E.g. my server should have quite a lot in the queue, since it is always on, > while my desktop computer (which is used only for several hours a day) should > only keep a few in the queue. > > The problem is that the settings for work/home/school aren't used by Boinc. > They all use the General preferences instead of the specific values I set. > > What am I doing wrong? > or > Is Boinc using the wrong preferences? > or > Is Seti not giving the correct preferences to Boinc?

9) Message boards : Number crunching : Progress doesn't grow evenly (Message 23925) Posted 9 Sep 2004 by Christopher Hauber Post: It's been a while since anyone has mentioned this problem. Maybe a good solution would just to make the time calculation logarithmic instead. Sort of "fudge" it. Still wouldn't be perfect, but would probably be closer. Especially if you tweaked the equation a little. > The algorithm for estimating the time to completion just doesn't work well for > SETI units. Ignore the estimate and the progress indicator, it's not linear. > >

10) Message boards : Number crunching : Crunching out of order (Message 23921) Posted 9 Sep 2004 by Christopher Hauber Post: Well I don't really know what the cause of it is, but I would imagine that the software is somehow misreading something. My computer is always connected to the internet so it connects to the server whenever it wants/needs to. Looks like when I wrote that I had about 75 unprocessed units which at a little under 3 hours per unit puts me at 225 hours which is just a little over 9 days I think. So I have about 9 days of work in theory. In reality that 9 will probably be more like 7-8. Maybe less I don't know for sure. In any event, it shouldn't be a problem on my computer unless it finishes processing the early deadline units after EVERY other workunit which I doubt. Even then it *may* be ok. It's only done 2 out of order. I've never really noticed it have this problem before so that's all I can really say. Something caused it, whether it be a bug or some really wierd fluke. And like I said, it probably won't effect me any, but I'm worried that if it really is a bug in the software it could happen again (and perhaps more severely so that it does effect me) and for others who really would be effected by it. > Last group of wu's I got was 20 with a deadline of 2004-09-15 and 20 with a > deadline of 2004-09-16. So far I haven't noticed any processing out of order, > and 10 of the wu's due the 15th are finished and reported. > > Only difference apparently between our system are the OS (and perhabs the type > of AMD CPU), I'm running Win2k pro SP3. > > Hmm. Hopefully your system remembers to take the deadlines into account. > > How many days have you set "Connect to network about every" to? (and how many > days do you estimate you have work for now?) > >

11) Message boards : Number crunching : Location not stored (Message 23916) Posted 9 Sep 2004 by Christopher Hauber Post: It's an option under SETI@Home preferences in your account. > How do you set the default to home? > > --Bill Z. > >

12) Message boards : Number crunching : Location not stored (Message 23841) Posted 8 Sep 2004 by Christopher Hauber Post: Yep. There really isn't much way to fix this except to do it manually. I mentioned something about this a while back. I later had a couple more minor issues with it (I think somehow I ended up with another copy of my computer and such which of course causes venue problems because the new "copy" is at the default venue and the new copy overrides the old one, so it can end up causing BOINC to use the preferences for the wrong venue again which requires a manual fix and so on. Kind of a nuisance. But since I currently have little reason to have the default venue be anything other than home, I set the default to home before I got too much into that cycle and only had to merge once. Kind of frustrating all the same though. Especially for those who are less computer inclined and would have trouble using the workaround. Chris > Here is a manual solution. Exit out of BOINC. Edit your client_state.xml > file. Search for the host_venue tag set. If your file was like mine, there > was no host_venue tagset. In that case add the following line > > school > > just before the last line which should be > > > > Once you have saved the changes, restart BOINC and you will see it now uses > the alternate prefs. This change will remain even after your projects contact > the server. > Thanks, > Michael >

13) Message boards : Number crunching : Crunching out of order (Message 23838) Posted 8 Sep 2004 by Christopher Hauber Post: I was checking the status of BOINC on my computer a few minutes ago and noticed that BOINC has been crunching SETI units out of order. It doesn't appear to be much of a problem for due dates since there aren't many with the early deadline left, but if it keeps doing that it could end up a problem or be a problem for others. Currently on the work tab I have 3 units showing they have been fully crunched (and turned in) with a deadline of 2004-09-15 @20:23:20. The one being crunched currently has the same deadline. In addition, there are 7 more workunits with the same deadline that have not yet been processed at all. There are 20 units with deadlines of 2004-09-21 @19:21:04 and 19:20:05, with 20 more due about 10 minutes later and another 10 due about 10 minutes after that. There are 20 more workunits that have deadlines of 2004-09-22 @ 02:17:28 and 02:17:29. Two of those workunits with the latest deadline (ending in 29 have already been processed and turned in even though though all but 4 other workunits are due before them of which a substantial amount are due an entire week before them. I'm running BOINC 4.05 on WinXP SP2 on an AMD system. My other computer hasn't been processing units apparently since I forgot to log it back in a few days ago after it was turned off so I don't know much more than this computer is telling me. Chris

14) Message boards : Number crunching : This is a new boinc project i think (Message 23560) Posted 8 Sep 2004 by Christopher Hauber Post: I don't really know what Globus is either but it has to do with grid computing and such. Looks like the they are using BOINC for biology applications. Clicking the names of the people on that site takes you to their personal info and provides a lot more information as to what it's all about. > thats what i got from it too andrew. none of the links do any thing they > either go to a pass word query or a forbidden access. it sounds like the > program integrates boinc with a few others. that will be cool because boinc > seems to take all the extra processor power and my gimps program (finds prime > numbers) doesnt get anything done. the problem is what globus is. i dont know > >

15) Message boards : Number crunching : We know it's broken, but why don't you tell us it is. (Message 23558) Posted 8 Sep 2004 by Christopher Hauber Post: Correct. 900K seems like a pretty good sum, but that's 300K a year. Educational research projects are expensive. Software development can be quite expensive. Heavy duty server hardware and software is expensive. Bandwidth is expensive. Support staff is expensive (even minimum wage staff since employees cost employers a lot more than the employee sees in wages even without benefits). This project has all of those. So it can be quite easy to tear through a $900k/3yr grant and funds from other sources. I think Ned breaks it down pretty well, it's just not a terribly large sum of money for what needs to be done. I would imagine David Anderson is running a reasonbly tight budget and is trying to manage it wisely so that funds will last until they are able to be renewed again. Even so, funding only gets you so far. Some things, once they happen, there is little that can be done to fix them, like having to transfer huge amounts of data around because the only machine you have with enough power to run things keeps failing. And that failing machine is still failing despite working many hours with the machine's support staff to fix it. Not really much that can be done to speed things up any so might as well wait. The funny thing about this though FatB, is that the post of mine you replied to had nothing to do with money or resources or anything of the sort. I just stated that the Dev Team had a lot more invested than most of us here, because they have directly spent countless manhours working on this project recently, whereas most people here have just crunched data or done alpha/beta testing. Anyway, I'm done now, thanks for the specifics on those numbers Ned. I think it helps put things in perspective a little bit. Chris > > OK Chris here are a couple of things you may like to mull over, for a > start go > > here http://www.nsf.gov/awardsearch/showAward.do?AwardNumber=0221529 and > see > > how much Boinc received in funding from the NSF, let me save you the > time > > $911,264 not bad for a 3 year award, which by the way runs out next > October, > > so I think they have had time to at least get it working a bit better > than > > they have. > > If we assume that the whole $900,000 goes to personnel, and not to computers, > bandwidth, tools, and etc. and assume that we're paying $40,000/year plus the > usual (office space, power and air conditioning, health insurance: costs > generally run about the same as salary) then the grant pays for 3 3/4 people > -- and nothing else. > > A comment was made that the money from NSF can't be spent on SETI, it is > restricted to BOINC. > > This article gives > the cost of bandwidth on the Berkeley campus as $300/megabit/month, and says > at that time they were using 25 megabits. If that's still true, that's > $7,500/month or $90,000 year. I can't believe bandwidth has gone down. That > leaves enough in the $250,000 Planetary Society grant to pay for a couple of > researchers. > > When you look at it that way, it isn't that much money folks. > >

16) Message boards : Number crunching : We know it's broken, but why don't you tell us it is. (Message 23419) Posted 7 Sep 2004 by Christopher Hauber Post: It should be obvious considering the time and effort they put into this project despite the constant stream of complaints that they take it quite seriously. They have a bigger investment in the project than most people around here do. > well it would be nice to know that seti admins take the project as seriously > as alot of the people here. > > There is no timezone on the internet. > >

17) Message boards : Number crunching : We know it's broken, but why don't you tell us it is. (Message 23417) Posted 7 Sep 2004 by Christopher Hauber Post: What enfuriated me were your "goldmine" and the "fortune and glory" comments. I knew that you were not speaking about the current state of the project, but rather an outcome based upon a VERY unlikely set of events. While there would be a lot of publicity about such a thing, and quite possibly some healthy grants issued for more research, the "goldmine" and "fortune and glory" comments to me were just over the edge and set forth at least an impression that the end goal and motivation of the project for the people running it and UCB was essentially greed rather than investigative science. As for SETI staff, they should probably be paid more. Some of them are paid I am sure. However I am not entirely sure if everyone working on it gets paid. I'm also not entirely sure if students can get credit for working on it. If they can recieve credit, then they are in effect PAYING to work on it rather than be paid. But salaries of employees of public institutions are generally public information and are subject to certain rules to help ensure that public money is used properly (kind of goes back to your "goldmine," kind of helps ensure that money is put back into research rather than into pockets). In any case, to some extent I agree that they are paid to provide workunits, but I think that is more an end result of their job than the actual job. Nitpicky perhaps, but perhaps relevant. They are really paid to write and fix the software and try to fix/maintain the hardware. In other words, try to get everything to work and keep it that way. The end result of that IS that workunits are issued to users, but the function to me (especially while things are still touch and go) just keep working on things to improve stability. I am not a "no problems here" user. Yes, there are problems. I still haven't downloaded units since they started being distributed a while ago, although until recently the scheduler was at least responding. Right not it appears to be non-responsive again and won't even let me upload. But oh well. That's how it goes sometimes. There isn't much I can do. Complaining about it won't cause it to miraculously resurect itself. The SETI team more than likely knows it's down, but it doesn't hurt to mention (not badger) it sometimes just in case they don't. I understand that. Some don't. And that is what bothers me most: the unfair senseless complaints and lack of understanding and patience. I do not stick my head in the sand. I do not think that criticism should even necessarily be avoided. It should however be constructive rather than destructive. It should also focus on what the problems are, not just that they exist. And when possible, it should also suggest alternatives, preferences for alternatives, etc. They are underfunded, underpaid, underequipped, etc. That has made things difficult for the team and is continuing to do so. However to some very limited extent, it may have been "self-inflicted." However, they had to make some decisions here and there and at times made the hard choice, specifically, the release date. It's hard to say with absolute certainty that they did the right thing all the time, and it is hard to say with absolute certainty that they made made costly mistakes in the decisions they made. It's all very gray. But what is done is done, and that cannot be changed. You should notice that in my posts I have not claimed that SETI is perfect said that legitimate complaints (like I mentioned above) should not be made. I have always supported posting problems people encounter and have posted some myself. What I HAVE spoken out against though are posts that flat out badmouth the efforts of the team and the project, those that demand and expect the project work perfectly, and those which make horribly inaccurate claims about the software or project. And for you to tell me that if I "cannot take any criticism of this project I would suggest you take your crunching elsewhere" indicates that you misread/misinterpreted what I have posted and have not read other posts by me. My primary post in this thread spoke out against the complaints about people having to wait longer for things to be fixed because of Labor Day -- not complaints in general. I suggest you pay more attention to what I write next time -- I generally try choose my words carefully so that I say what I mean and mean what I say. Chris > Christopher, maybe you need to calm down then. I was responding to the issue > of money and payment which was raised by Nedd. I am not suggesting the project > goal is to make money but it will be the outcome (I believe). If you cannot > take any crtiscism of this project I would suggest you take your crunching > elsewhere. I do however happen to agree entirely with your views on the Seti > team having a holiday. > > I believe the Seti staff are paid - I'm not suggesting they are getting rich > or making big bucks but they are paid. This may be a not for profit but that > doesnt mean unpaid volunteers (actually it does you and me). I can only > presume they get paid to supply you and me with work units otherwise the > project and their jobs would cease to exist. This is not meant as a critiscism > just fact. > > What I do not agree with is the "no problems here attitude". There are many > problems including insuficient funding, lack of resources, lack of information > being given etc etc. Some of those problems the Seti team can do nothing about > - some they can. > > If you want to stick your head in the sand and say "all OK, how dare you > criticise" then so be it. If you want to defend the indefensible so be it. I > believe it is not that black and white. > > I am not suggesting money is the goal or driving force however, if and when, > somebody will make a lot of money IMHO. > > The Seti team are clearly doing a difficult job at a difficult time however at > least part of those difficulties are self inflicted. > > Neil

18) Message boards : Number crunching : We know it's broken, but why don't you tell us it is. (Message 23405) Posted 7 Sep 2004 by Christopher Hauber Post: Yes they would. But most faculty have keys to their buildings and labs. Same with GTAs. Some campuses also allow students into select buildings, labs, and rooms depending on their major and classes 24/7 electronically using their student IDs. I do not know if UCB provides that or not, but nonetheless, Labor Day or no, arrangements for students and faculty to use university facilities can generally be made but is stritcly voluntary. That is why I made mention of it. It may be possible, but should not be expected. Chris > Actually the buildings on the campus would be locked so unless they had the > key to get into the building itself they wouldnt have access. I dont know how > UCB is set up but the faculity may not have building keys as they would office > keys. Being so close to the start of classes I doubt even the library would > be open. > >

19) Message boards : Number crunching : ETBOL...! (Message 23077) Posted 6 Sep 2004 by Christopher Hauber Post: You shouldn't need to contact the schedulers to upload units. Only to download. Read my post in this thread: http://setiweb.ssl.berkeley.edu/forum_thread.php?id=3813 > ETBOL > > Estimated time back online! > > I have been waiting for a few days to up load my work packages, but all I get > is > “No schedulers responded” … > I have no idea when or how long I have to wait! > > So how about an ETBOL so I know when I can turn on my pc…. > > Martin… > >

20) Message boards : Number crunching : We know it's broken, but why don't you tell us it is. (Message 23076) Posted 6 Sep 2004 by Christopher Hauber Post: Yes it is an international project. I've had to explain that to people on numerous occasions for various reasons. That is why I took special care to specifically say that the project is in the US, and Labor Day is a US holiday, not celebrated by most other countries (if any). But this is a case where the international aspect of the project is set aside for a moment. The international (and stateside for that matter) population supporting the project needs to understand that because UCB is a public institution in the State of California, in the United states, its faculty and students are given Monday off. But as an educational institution where some like to take advantage of the time to catch up on stuff and work on things, the option for people to use the facilities (especially faculty) is generally there. But that is strictly voluntary, and most students take the extended break as an opportunity to go home. I don't care that SETI benefits from 24/7/365.25 contributions from all over the world, it doesn't change the fact that there are local holidays and such are going to effect the ability of the Dev Team to work on the project. It's just a rare special case with this project when international aspect of the project ceases to be of an important influence, and rightfully so. > To state the obvious, Christopher, this is an International, Global project. > > SETI benefits from World-Wide 24 hour, 24/7, 365 day/year contributions. > > I (we) respect the limitations of the development/support team, and the time > investments they make. > > But please remember that many, certainly in Europe (a fair percentage of the > world total contributors), find ourselves 23 hours behind the events in > California - and to have that pushed to 48 or 72 hours is - to say the least > frustrating. > > To the Dev team, enjoy your holiday, and lets look forward to a really > sucessful and exciting fall :-) >

Next 20

 
©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.