Posts by Bill Michael

1) Message boards : BOINC Manager : BOINC slowdown on Mac OSX (Message 2637)
Posted 16 Jan 2006 by Bill Michael
Post:
I'm lost... all I can do is throw some suggestions. If you have more than one user set up, log in as the other user, see if it's running there. Use software update and make sure you're up-to-date. Dismount the iDisk. Shut down, turn off any external drives and reboot. Unplug network cable. Remove everything from your list of startup items and reboot, then put them back one at a time. Run a virus scan. Make sure there's not a corrupted CD/DVD in the drive, or some USB/Firewire storage device attached that has a problem (digital camera, iPod, etc.).

As you can tell, I'm just guessing on all of it! :-(
2) Message boards : BOINC Manager : BOINC slowdown on Mac OSX (Message 2623)
Posted 16 Jan 2006 by Bill Michael
Post:
Spotlight indexing? iDisk syncing? Either of those would continue after a reboot. Activity Monitor _should_ give _some_ indication of what's taking the resources, even if it's a system process. You may have to select "all processes" from the drop-down.
3) Message boards : BOINC Manager : How do I speed Boinc up? (Message 2618)
Posted 15 Jan 2006 by Bill Michael
Post:
Wow, didn't realize a 3700 would run that fast. I'm running an AMD 3000+ 64bit 939 and it takes me a little over a hour to complete a unit. So I guess it is the processor after all.


And Crunch3R's SETI app! If you already have a 3000... if your motherboard supports it, I would bet you could buy a dual-core CPU, either a 3800 or a 4400, and get a _far_ bigger improvement for your $ than you'd get by adding a whole new system for the same price as the dual-core CPU alone.

Obviously if you can _add_ an X2 system, and keep the 3000, that's even better; but if the budget says you're going to wind up with a single-core second box, I wouldn't do it; you're putting too much of the money into other things besides the CPU.
4) Message boards : BOINC Manager : How do I speed Boinc up? (Message 2615)
Posted 15 Jan 2006 by Bill Michael
Post:
I don't know your budget; mine was fairly low; but I did this myself just a few months ago. I wound up with an ASUS motherboard (socket 939) and an AMD 3700+ "San Diego", a "gaming" case (lots of fans as I knew I would be overclocking, but not "lots of fans where they look good" - actually where they DO some good), a $50 PCIE video card, 2 256MB sticks of RAM (which I already had; should have spent a bit more and gotten faster RAM, again b/c of the oc'ing, I can only get to 2530), small but fast SATA hard drive, used junk parts for CD and floppy.

You will need at least 512MB of RAM, and you don't want to get the "budget" stuff; if you're going to overclock, then you'll need the "really good" stuff. Other than that, the only thing that really matters is the CPU, and that the motherboard is "new" enough and "good" enough to _truly_ get the most out of the CPU. Video only matters when you are watching the graphics. You _don't_ want to scrimp on the power supply; the one that came with my case fried after two months, taking out the video card.

The only thing I would change about my system, other than better RAM, is that if I had the $$, I would have put an AMD X2 4400+ in it instead of the single-core 3700. If you want to see mine on SETI, it's here - it does a SETI result in about 35 minutes. Not bad for around $700!
5) Message boards : BOINC Manager : feature request: a suspend/resume button on transfers tab (Message 2607)
Posted 15 Jan 2006 by Bill Michael
Post:
This has been proposed on the developer's mailing list... "Network activity suspended" does ALL projects, and there needs to be a way to suspend individual transfers. No idea when this will be added.
6) Message boards : BOINC Manager : To those that use Macs, or Linux, or OTHER (Message 2596)
Posted 14 Jan 2006 by Bill Michael
Post:
Actually, I would bet that 75% or more of the questions asked on the boards are from Windows users. The Mac-specific help areas that I monitor are _much_ less active than the Windows ones, probably even slightly less than the % of machines would indicate. And in general, Linux folks are far more knowledgeable about their systems and how to make things work. As far as BOINC itself goes, I run it on both Mac and Windows, and they both seem to be almost identical. Well, I'll take that back - my Windows box has the "can't connect to localhost" problem on startup at times.
7) Message boards : BOINC Manager : Activities suspended under Work tab (Message 2573)
Posted 13 Jan 2006 by Bill Michael
Post:
Check the Commands menu too; step one, set that to "run always" and make sure the problem disappears. Then try "based on preferences", and if it stops work, you know it's something in your webpage prefs...
8) Message boards : BOINC client : Does BOINC affect a network? (Message 2567)
Posted 13 Jan 2006 by Bill Michael
Post:
I need an official answer for a network admin and a knowledgable person on the board (whilst helpful as I stated and offered my gratitude) won't be enough for them.


You'll need to post on the SETI boards then, or use email; and if the boards you'll have to decide what "official" is, if a "volunteer developer" or "volunteer tester" or "forum moderator" is good enough, or if it has to be a "project scientist" or something... Be clear up front in your posting, who you require an answer from, and why; Paul's problem with your response _is_ understandable, because you didn't explain _why_ you couldn't just take John's answer. Once we know some "suit" is involved, well, we've all been there. But "need concrete facts" is pretty dismissive of what John gave you, and "or someone who has qualifications" is pretty dismissive of him personally; they _were_ the facts, as concrete as you're going to get, and he probably actually KNOWS more about the issue than someone "on staff" - you just have a particular, bureaucratic-not-technical, requirement.
9) Message boards : BOINC client : running BOINC client without taskbar entry (Message 2560)
Posted 13 Jan 2006 by Bill Michael
Post:
BOINC Manager automatically runs the BOINC daemon... if it isn't, I would reinstall BOINC.
10) Message boards : BOINC Manager : I miss the old SETI@Home (Message 2517)
Posted 11 Jan 2006 by Bill Michael
Post:
I don't understand how SETI Classic could work without open ports. It had to "call home" to get work and to upload finished results, the same as BOINC does... BOINC may require _more_ particular ports be open, ie; 1043 for internal communication _IF_ you want to run BOINC Manager, instead of just port 80.
11) Message boards : BOINC client : can we have pre release versions (Message 2499)
Posted 10 Jan 2006 by Bill Michael
Post:
I just looked, and I see 5.2.14 (bad version), 5.2.15 (in test), 5.3.2 and 5.3.6 (in development). Are you looking at the "public" download site rather than the "full" download site?

http://boinc.berkeley.edu/download.php?dev=1 for currently in-test versions, or

http://boinc.berkeley.edu/dl/ for everything...

DISCLAIMER: Using ANY of these links is done "at your own risk", same as if you downloaded the source and compiled it yourself; any version other than the "currently recommended" one may cause you to lose results, wipe your hard drive, and cause tooth decay.
12) Message boards : BOINC client : Scheduling Needs Project Out-of-work Flag Kept in Client (Message 2498)
Posted 10 Jan 2006 by Bill Michael
Post:
It been my experience that if a project doesn't have work, the duration is for a long time, versus 5 minutes or so...


My experience is exactly opposite - Predictor would have no work for my platform, then next connection would. Rosetta gets some overload condition where it will say "no work", but retrying gets work. SETI has been running "on the edge" between having work and not, with people grabbing it as fast as they can generate it. I've never attached to LHC, they have no Mac client.

But to your specific problem; if you have three projects, and there is no work ON YOUR HOST for two of them, then regardless of any "out of work" signals from those two projects, the Client should download your full cache setting worth of work for the third project. If it is not doing that, then it may be a bug, rather than a need for a new feature.

Work fetch is controlled by long term debt. LTD is not supposed to be affected while a project is out of work, but it does slowly drift. This is why the standard recommendation is to suspend such projects, as that completely removes them from the LTD calculation, or even to occasionally "reset" them (which resets their debt to zero). If you let a project be active for a long period while it has no work, the debt for that project will climb, which (because the sum of all LTDs is zero) will affect the LTD of your _active_ projects, and therefore affect work fetch. Did you note the LTDs for the projects before detaching? If they were the cause, that's a _known_ bug with known workarounds, as mentioned. If not, we could have tried to identify the problem and see if it was something new.

There is no currently available _reliable_ way to detect a dialup user. The upload/download limits are not _required_ to be set; I would bet 99% of dialup users say "no limit".
13) Message boards : BOINC Manager : I miss the old SETI@Home (Message 2494)
Posted 10 Jan 2006 by Bill Michael
Post:
Do you know how to get BOINC working through a strict firewall in Windows. I cannot open any ports (not up to me), the proxy settings are greyed out.


If the owner of the PC will not allow you to run BOINC (which means opening ports), you are violating the terms and conditions by running it. So - don't run it on that PC.
14) Message boards : BOINC Manager : How to delete an aborted result? (Message 2481)
Posted 10 Jan 2006 by Bill Michael
Post:
Is there a way (from BOINC Mgr) to delete a result in status "Aborted by user" short of resetting the project? I have an aborted CPDN result sitting on my machine, taking up several hundred megabytes of disk space. I have 2 other CPDN results crunching along, and I'd hate to dump them with a reset.


Set "no new work" for CPDN. In the Work tab, "suspend" one of the currently-running CPDN results. It should then try to run the aborted one, which will cause it to be sent. "Resume" the result you suspended, and all should be well.

EDIT:: Les's right, this will get rid of the _result_, but not the disk space. For that, AFTER you've gotten the result sent, you'll still need to delete the files.
15) Message boards : BOINC client : Feature Request: Project Specific Cache Size (Message 2479)
Posted 10 Jan 2006 by Bill Michael
Post:
Actually, a better fix for the EDF problem is already "on the list" - to take into account the resource share when doing work fetch. Then when it gets "x" days work, it'll get 90% from your 90% project and 10% from your 10% project.

I'd still like to be able to set the cache per project though... it's just that that is a much "larger" change to the code, so it's unlikely we'll see it soon.
16) Message boards : BOINC client : Scheduling Needs Project Out-of-work Flag Kept in Client (Message 2476)
Posted 10 Jan 2006 by Bill Michael
Post:
Instead of detaching, the normal approach is to suspend any projects that don't have work; that accomplishes the same thing, but it's easier to hit "resume" than it is to reattach, and you can resume every third or fourth time you connect, just to see if there is work.

The problem with a project being "out of work" is that BOINC doesn't know when it's suddenly going to _have_ work - that could be five minutes from now. If there was a way to tell when someone is on dialup, it would be easier to deal with this, but all attempts to get "dialup yes/no" or separate settings for connect-every and cache, have been refused so far.

Which version of BOINC are you running? There are complaints about current versions (5.2.13) _not_ following resource shares during work fetch, and getting too _much_ work.
17) Message boards : BOINC Manager : Automating Project Attachment on Remote Machines (Message 2470)
Posted 9 Jan 2006 by Bill Michael
Post:
I'm also assuming I can figure out what the necessary xml files are by poking around the messageboards here and/or at the project sites?


No - you'll need to look at the xml files in the already-attached working folder. My guess is that you'll need _all_ of them, but only the one time. You DO NOT want to have any work on the computer that you copy the files from, it is these files that hold the list of work to be done; if you copy them from a system that has work, the new host will error out trying to find the missing work.
18) Message boards : BOINC Manager : BOINC crashed (Message 2447)
Posted 8 Jan 2006 by Bill Michael
Post:
Ok, now I am stumped. Bill, any thoughts?


None... never seen this one, no ideas.
19) Message boards : BOINC Manager : Automating Project Attachment on Remote Machines (Message 2444)
Posted 7 Jan 2006 by Bill Michael
Post:
Thank you for the response. Out of curiosity, is there a "complex" way?


Sure - use the installer on all the machines, then copy in just the necessary xml files for each project...
20) Message boards : BOINC Manager : Automating Project Attachment on Remote Machines (Message 2430)
Posted 6 Jan 2006 by Bill Michael
Post:
The simple way is to set up a BOINC installation that is going to be copied to the other computers; attach to whatever projects you like, set "no new work" and run "dry"; then instead of downloading a new BOINC installer for each remote machine, install a copy of the one with everything already set up.


Next 20

Copyright © 2021 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.