Posts by ChertseyAl

Info	Message
81) Message boards : Projects : FightMalaria@Home Message 44914 Posted 15 Jul 2012 by ChertseyAl	Be careful with this one. Although there is often a lot of work available, it seems the DB gets wiped frequently and unless you update after every WU you may waste a lot of crunching time. Been there, done that :) However, looks like an interesting project. Let's hope they get the message boards up ASAP so that we can all find out more and start contributing! Cheers, Al.
82) Message boards : Projects : Rioja Science Message 44570 Posted 20 Jun 2012 by ChertseyAl	Project is probably using coarse HR, which matches Windows, Linux, and Mac with the same OS. I think that's true as I can't find any recent work where my XP machines have been paired with linux. Which is nice :) Still getting a few validation problems, but maybe only 1 WU per day, so not that worried. HOWEVER ... It would be nice to lift the limit on WUs in progress per host/CPU/whatever to a sensible level. My machines hit the limit, keep asking for work, back off for hours, and as a result machines are often dry :( Cheers, Al.
83) Message boards : Projects : Rioja Science Message 44439 Posted 9 Jun 2012 by ChertseyAl	The feeder hasn't been running for a while now. I have work to report, there is work available to get, but someone needs to give the feeder a kick please :) Cheers, Al.
84) Message boards : Projects : Rioja Science Message 44282 Posted 24 May 2012 by ChertseyAl	Perhaps this thread should be locked. I disagree. Fair enough. Meanwhile, a few problems have come to light: "Completed, validation inconclusive" - This is still happening with linux v Windows, but now more worryingly seems to happen between Win XP32 and Win7 64. Also, a 'tie-break' WU is not being sent, so the tasks presumably never get resolved. The tasks don't relinquish control once running. Once they start, they run to completion, blocking other projects. Checkpointing doesn't work, or isn't present (probably the same issue as above, it simply isn't implemented). Prior to a power failure I had some WUs running solidly for 5 hours. On restart they started from zero again. For some reason, the IR is set to 3 although the MQ is 2. I thought this was generally frowned upon as a waste of resourses. However, on the bright side, only 2 replications are actually sent out, the third remains as 'unsent' (see validation issue above, probably related). I'd suggest it's pretty important to sort out the validation issues as by my estimation about 10% of work is being wasted (seems to tie up with the 90/10 win/linux split that someone quoted someplace). Secondly, as the WUs are a decent length, checkpointing would be nice :) Cheers, Al.
85) Message boards : Projects : Rioja Science Message 44277 Posted 23 May 2012 by ChertseyAl	Perhaps this thread should be locked. The project's admins really must open their message boards. There are so many issues that could be resolved with helpful crunchers if only there was a forum (sic) to communicate. Looks like the WU length has increased, as suggested, by 100x. Looks like all of the win v linux WU validation problem WUS have been wiped (well, gee, thanks for that). Looks like there's still a limit on concurrent WUs. FÃ©lix, get those message boards open. Now. Before we all lose interest in what is potentially an interesting project. Cheers, Al. p.s. Someone please lock this thread!
86) Message boards : Projects : Rioja Science Message 44163 Posted 18 May 2012 by ChertseyAl	FWIW, there seems to be a problem validating linux versus windows machines. All of the 100+ pending WUs I've checked so far have been linux against windows. Not an uncommon problem. Also, stats have not been updated since 16th May, but nonetheless I'm grateful that you are at least trying to export them :) Oh, and the limit on WUs in progress is proving a real hindrance to getting any work done. Suggest a minimum of 2 WUs per core, but ideally a chance to get an hour or mores worth of work cached would help :) If the WUs took a bit (lot!) longer it would make the project easier to crunch, but I don't know how practical that is for you to fix. 100 times more work per WU would be nice. Sadly still no message boards :( Cheers, Al. p.s. Sorry for cluttering up the BOINC dev boards :)
87) Message boards : Projects : Rioja Science Message 44154 Posted 15 May 2012 by ChertseyAl	Oh...And please state that it's okay to import and publish project statistics. As stats aren't being exported it would be quite hard for any stats site to import them :) Anyway, we shouldn't be cluttering up the BOINC forums with project-specific stuff - Let's hope the project opens it's forums soon :) Cheers, Al.
88) Message boards : Projects : Rioja Science Message 44056 Posted 8 May 2012 by ChertseyAl	IÂ´m working to fix the problem, sorry for the inconvenience. Hi FÃ©lix, Working for me now. Maybe time to open up the message boards on the project site to avoid traffic here? Cheers, Al.
89) Message boards : Projects : Rioja Science Message 43966 Posted 4 May 2012 by ChertseyAl	All work units failed for me on Intel XP32. No message boards, no way to help them fix the problem. Oh well. Cheers, Al.
90) Message boards : Questions and problems : Plot showing CPU Life vs. Usage Message 43205 Posted 29 Mar 2012 by ChertseyAl	FWIW, I've never, ever had a CPU fail. As you mention 'computer lifespan', maybe you want to know about other components? If, so, here's my list of component failures, most common first, with my experience on non-BOINC machines added: 1) PSUs - These fail regularly on my BOINC machines. Lifetime is probably around 2 or 3 years. I can only recall one failing on a machine not running BOINC. 2) CPU fans - Less common now, but the bearings wear out leading to overheating and either crashes or BIOS shutdowns. Again, only happened once on a non-BOINC machine, but that did get heavy use in other ways. I know some CPUs did pop if the cooling fan failed, so maybe that's the factor in CPU failure running BOINC. Cheap bearings sieze due to running at 100% 24/7. 3) HDDs - Have lost of few of these over the years. Nearly always newer SATA ones, which is interesting/worrying. Non-BOINC, probably only one standard drive, but laptop drives seem flaky generally. I don't do BOINC on laptops, apart from one hateful Vista lappy that I'm trying to kill ;) 4) Motherboards - Have scrapped a few machines due to dried out caps. Not worth fixing when the whole board is covered in popped caps! Possibly not BOINC related, but the higher case temperatures probably don't help. Oh, I only use cheap (or preferably free!) machines that have been abused in offices or universities for a couple of years before I get them, so I always expect early failures :) Al.
91) Message boards : Projects : Donate@home Message 42759 Posted 26 Feb 2012 by ChertseyAl	Hehe. Pulls up chair, opens popcorn ... ;) Al.
92) Message boards : Questions and problems : Repair install on Ubuntu Message 42598 Posted 16 Feb 2012 by ChertseyAl	FWIW, just shutting down BOINC and reinstalling repairs the install and leaves all existing projects and WUs in place. Which is nice :) Al.
93) Message boards : Questions and problems : task management is useless Message 42533 Posted 11 Feb 2012 by ChertseyAl	basically the scheduling of tasks is bonkers It does indeed seem very strange sometimes. Rather than fight it, and just be told a million times that you are wrong and that you shouldn't micromanage etc etc, why not just set a zero cache and tiny 'extra work' time (e.g. 0 and 0.01) so that your machine gets just one WU per core that runs to completion and then requests new work? This assumes that you have a 24/7 net connection of course! Cheers, Al.
94) Message boards : Questions and problems : Repair install on Ubuntu Message 42532 Posted 11 Feb 2012 by ChertseyAl	I've got a couple of Ubuntu 8.04 hosts running 5.10.45 that have 'lost' a project or two (can't connect, can't attach etc). When this happens on my XP machines I just run the installer and select 'repair' and all of the broken projects come back to life and keep all of my existing projects and WUs :) No idea how to do the same in Ubuntu though. Ideas? Cheers, Al. p.s. Not interested in upgrading the OS or the BOINC version thanks!
95) Message boards : Questions and problems : Ubuntu Install Message 42531 Posted 11 Feb 2012 by ChertseyAl	Last time I did it I typed the following into a terminal window: sudo apt-get install boinc-client boinc-manager You might want/need to substitute 'aptitude' for 'apt-get'. I Am Not A Linux Expert - hopefully someone else will give yoy the full rundown :) Al.
96) Message boards : BOINC Manager : EXPLICIT installation instructions please Message 42436 Posted 3 Feb 2012 by ChertseyAl	I did a quick install the other day. I just opened a terminal window and typed: sudo apt-get install boinc-client boinc-manager It installed OK (but it was version 5.10.45, old version of Ubuntu), I opened BOINC Manager, connected to a project with my email address and password for that project and I was up and running. I don't think I've ever used the command line to set up BOINC on any of my Linux boxes. Perhaps you are trying to do something more complicated than me. I am not a Linux expert BTW! Al.
97) Message boards : Projects : Projects for not often turned on laptop/pc Message 41340 Posted 27 Nov 2011 by ChertseyAl	Some projects with very small workunits or really long deadlines? First, set a zero day cache with a tiny amount of additional work buffer to keep the number of WUs to a minium, ideal 1 WU downloaded per core. But you knew that already :) For projects, two approaches: Long deadlines: Climate prediction. You'll never complete a WU, but the project 'trickles up' so you'll be doing something useful. Don't know if BOINC will kill the WU once it's gone over deadline and download a new one, but CPDN don't seem to worry about deadlines anyway. Small WUs: Primegrid subproject Proth Prime Search LLR would be ideal. About 20 minutes per WU. Al.
98) Message boards : Questions and problems : Does BOINC run better on Windows compared to Linux? Message 40938 Posted 3 Nov 2011 by ChertseyAl	Recently a fellow cruncher stated that BOINC was optimized for Windows and would be less efficient on Linux. Is this true? A while a ago I benchmarked a PC across as many projects as possible with a clean Win XP install. I then made a clean Ubuntu install and repeated the benchmarks. On average Ubuntu was 33% slower than XP. There were exceptions, 2 projects in particular were twice as fast. Some were notably agonisingly slow (Spinhenge, take a bow!). Nothing to do with BOINC though. Just the apps from the projects, how they were compiled etc etc. This was a while ago, can't even find the spreadsheet that I made as I was doing the tests, things have moved on since then anyway. I only run a couple of linux boxes these days just to try projects that are linux only. Rest of the time I just let them crunch a couple of maths projects that they seem to do quite well. Why not make your machines dual-boot or run VMs and try it for yourself? Would be both interesting and useful for all of us :) Al.
99) Message boards : Projects : EDGI Demo Project Message 40878 Posted 28 Oct 2011 by ChertseyAl	How do you find all these new, obscure projects? Actually, just googling for "is a research project that uses Internet-connected computers to do research in" and digging deep will reveal all sorts of startup projects. Been there, done that, still do it sometimes :) Al.
100) Message boards : Projects : SKA Distributed Project Perth :Western Australia Message 40421 Posted 28 Sep 2011 by ChertseyAl	http://www.boinc-australia.net/ ??? Would have thought http://www.theskynet.org/index was more relevant ;) Al.

Copyright © 2026 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.