Posts by Paul D. Buck

1) Message boards : BOINC client : Boinc V6.10.4 bug fix annoyance (Message 27190)
Posted 9 Sep 2009 by Paul D. Buck
Post:
On the Alpha mailing list which I just finished catching up on ... they introduced a bug in Feb 2009 by turning off something in DCF ... so, when you start 6.10.4 the DCF goes way low and then recovers ...

From the list (Dr. Anderson):


The key here is "on_frac 0.027".
The code that updates on_frac (fraction of time BOINC is running)
was accidentally disabled in Feb. 2009.
6.10.4 fixes this, but it thinks you haven't run BOINC since Feb 2009,
so your on_frac becomes small.
That makes BOINC give long estimates for job completion.
(Eventually on_frac will return to its proper value,
but will take a while).

I'll put a cap on the on_frac decay to deal with this.
In the meantime, you can edit your client_state.xml,
look for <on_frac>, and change it back to 1 (or whatever it should be).


They mention updating the value in the xml files ... the long and short of it is that any upgrade to 6.10.x with x>3 is going to have this issue ... until you work past it ...
2) Message boards : Questions and problems : Not getting new WU's for CPU projects after upgrade to 6.6.20 (Message 24489)
Posted 22 Apr 2009 by Paul D. Buck
Post:
No replies on the mailing list yet, but I'm waiting for fuller particulars from server administrators - I've asked at Einstein and CPDN, two projects I know well which have had recent server overload problems, if they feel that they're suffering from what I'm calling "content free" schedulewr RPC calls from clients. Thyme Lawn has guided Milo in what to look for in the server logs - Jord has access to the area where the question was put, and where any answer will appear. None yet.

I am pretty sure you are correct as this being a problem. Might also want to consider trying to get the info from MW too ... my bogus GPU calls to them may be part of the problem that they are seeing also.

The mind set is that if the cost appears low there is no reason to avoid doing stupid things over and over and over again. I have still to hear why we want to reschedule CPUs once every 60 seconds much less more often than that. Yet that at its root what is causing the issue we have been trying to get them to recognize ...
3) Message boards : Questions and problems : Not getting new WU's for CPU projects after upgrade to 6.6.20 (Message 24485)
Posted 22 Apr 2009 by Paul D. Buck
Post:
There's at least one bug still in work fetch, which can look like the problem reported here. I've reported it as a v6.6.23 bug, but the relevent section of code hasn't changed since v6.6.20

I don't recall seeing this on the mailing lists. Or is it being talked about elsewhere.

Not that I consider most of the discussion responsive on the mailing lists.
4) Message boards : Web interfaces : Computers not hidden but boinc.berkeley.edu says they are? (Message 2990)
Posted 8 Feb 2006 by Paul D. Buck
Post:
Ah, Ha!

A simple explanation ... thanks James! :)
5) Message boards : Web interfaces : Computers not hidden but boinc.berkeley.edu says they are? (Message 2987)
Posted 8 Feb 2006 by Paul D. Buck
Post:
Urgh...

Ok, no idea why ...

SIMAP does not use the standard forums so I am not always sure if I trust the reports ... :)

Though I have a BURP account, I never did any work there and don't pay any attention to the project as they don't seem to be coming out of testing ... or at least not anytime soon ...

Good question and I have no answer ...
6) Message boards : Web interfaces : Computers not hidden but boinc.berkeley.edu says they are? (Message 2980)
Posted 8 Feb 2006 by Paul D. Buck
Post:
Pam,

When did you show your computers?

If the UCB site has not updated or the BURP site has not exported the data, that is the problem. So, if BURP is not making the exports then the UCB site cannot update.
7) Message boards : BOINC client : ../windows/projects/setiathome.berkeley.edu removal? (Message 2963)
Posted 5 Feb 2006 by Paul D. Buck
Post:
Posting the same question in multiple threads will make it harder to get good help. If you look I answered this question in the other forum. So, mark one of these as closed and only use one thread.
8) Message boards : BOINC Manager : ../windows/projects/setiathome.berkeley.edu removal (Message 2962)
Posted 5 Feb 2006 by Paul D. Buck
Post:
There is a thread on the SETI@Home board that talks about this issue.

Hopefully someone smarter than I will be able to tell you how you might detect the activation of the install.

Easiest to check is the "Startup folder" to see if there is a shortcut to a program that should not start.

Next would be to use TweakUI" to see what applications are autostart. Again, I don't know all of MS's names so I could not advise. BUt, if you could make those first two checks it would help them, especially if youpost the results. I know that there is a high amout of interest in finding out how this is happening.
9) Message boards : BOINC Manager : Multiple seti, einstien and climatecontrol projects running simultaneously (Message 2904)
Posted 1 Feb 2006 by Paul D. Buck
Post:
CPDn work is usually issued with a one year deadlin. So, that should not be a problem on a reasonably fast machine. Otherwise. just abort one ... the sooner the better. I usually set CPDN to no new work just in case.

There have been some issues in recent times with CPDN downloads mostly because we are just now really getting a handle on the issues with a wide variety of work unit sizes, deadlines, etc.

The good news is that we ARE making progress and the BOINC Client Software is getting smarter all of the time.
10) Message boards : BOINC Manager : Mysterious error 500 still with me (Message 2892)
Posted 31 Jan 2006 by Paul D. Buck
Post:
I have some of the technical details about this error in the wiki now. Tigher gave me a start that I am still formatting (bear with me), but you can see a little about what we know and are trying to learn.

For the networking mavens, if you have constructive suggestions I am all ears ...

See here There is new summary information at the top, and the new block at the bottom.

One thing I have been wondering is if we should split some of it out or not ...
11) Message boards : BOINC client : Scheduler and CPDN (Message 2871)
Posted 30 Jan 2006 by Paul D. Buck
Post:
YOu can get into a couple of other problems with the various modes and CPDN (and a few other combinations). The good news is that John has once again figured out what most of them are and we do have a lightly better CPU scheduler and Work Scheduler in the mill.

The issue has been that this is not an easy component to work on. It has gone through at least 3 major generations with new projects, like CPDN, causing possible issues.

Heck, I have the opposite problem ... I have a CPDN work unit I can't keep IN EDF mode so it will finish on time ... :)
12) Message boards : BOINC Manager : Won't login (Message 2870)
Posted 30 Jan 2006 by Paul D. Buck
Post:
The later versions of BOINC want to connect with a username and password. This is causing no end of confusion for those of us that are used to the account keys (which I still use). You are supposed to be able to use the account key as your password with no username in the attach wizard. I have been remiss and have been using the BOINC View attach so I don't have as much experience as I would like anymore with the new system.

The other thing to do is to goto the Rosetta@Home and in the your account page create a password for the account and then use that in the wizard.
13) Message boards : BOINC Manager : Newbie Questions (Message 2803)
Posted 27 Jan 2006 by Paul D. Buck
Post:
1) BOINC uses "idle" time, so the processors will run at 100% at all times. When you do something else BOINC will "back out of the way and let the other thing take over. when there is idle time again BOINC and the sicence applications take up the "free" CPU time.

More on this and other things can be found in the Wiki, start with CPU and move on from there. Read the FAQ, do "Credit" and other things for fun too ...

2) You cannot run both at the same time. I do not know why it does not compress and mount correctly for you, it did for me. What version of OS-X? do you have an up to date decompressor? Did you try to decompress with BOINC Menubar turned off?

3) Yes

Edit, CPDN does have a Mac version. It always crashes on mine, but others have run them successfully on OS-X. For a more complete list of projects see the wiki ... :)

==== edit #2
What is the URL of the project you are trying to join?

CPDN is http://climateapps2.oucs.ox.ac.uk/cpdnboinc/
14) Message boards : BOINC client : Irritating (Message 2762)
Posted 25 Jan 2006 by Paul D. Buck
Post:
The wiki is still your friend.
15) Message boards : BOINC Manager : I'm gone. (Message 2761)
Posted 25 Jan 2006 by Paul D. Buck
Post:
If you look at the other projects they have similar patterns ...

which could mean that the migration month was hard on everyone, and some left because they could not get BOINC to work.

Or, the holiday month could have put a damper on participation as people were doing other things.

Some went to LHC as they had work so they moved their computers over while there was work.

Also, some left SAH for the simple reason that they knew this was going to be a rough time and they went to more stable projects but fully intend to return.

You can spin the data however you like. All it shows is that there was a "blip" in the data.

The general trend is still up, with the exception of Predictor@Home which seems to be bleeding people a little faster than they have been gaining them ... and there are always those that try a DC project and leave.

I don't know if there are really good stats for Classic of active vs. total participants, but, my recollection is that the numbers were worse there ... 5.5 million total, less than 500,000 active which in my book is worse than the 50/50 ration which we are close to ...

Again, it could be better ...
16) Message boards : BOINC Manager : How do I speed Boinc up? (Message 2692)
Posted 21 Jan 2006 by Paul D. Buck
Post:
There are discussions of these issues in the Wiki under selecting a BOINC Powered Project. However, since there is a wide variation in actual run times you do have to test each project because one may be more appropriate for your computer and another to a person with a slightly different configuration.
17) Message boards : BOINC client : Confused About Report Deadlines vs. Credit (Message 2691)
Posted 21 Jan 2006 by Paul D. Buck
Post:
Deb,

I am not sure why you are having these problems as there are others that are still successfully doing SETI@Home work on computers similar to yours. 70 hours to complete is not bad with a two week deadline. The only issue I can see is that you must not be running the computer as much or there is some other issue.

The SETI@Home science application that BOINC is currently using is virtually identical to the one that was being used in Classic days. Better, there are optimized versions that run considerably faster that are available ...

Without more details it is difficult to help you, and since your problem is more with SETI@Home you would have better luck posting a question in the Q&P forum in the SETI@Home web site where you can make your computers visible and show the actual work performed. Tony over there is very good at helping people.

The use of the deadline is because work is issued and then validated on return. In Classic, well, these last few months have been taken up with validation of the work that had been submitted over some considerable period of time.
18) Message boards : BOINC Manager : How to attach to a project in alpha stage ? (Message 2674)
Posted 20 Jan 2006 by Paul D. Buck
Post:
It is an older project and though they say they are alive, and I have an older account I have not been able to re-attach. I notified the contacts at UCB, but, aside from saying that they are still alive ... no good word as far as I can see ...

Because they do not have the correct "ID" files the newer BOINC Clients cannot attach ... trying through BOINC View, which seems to be more tolerant, the project is listed as "Down" (still) ...

So ...

This is one of the reasons I moved Crash Collection and Resource Measurement to inactive ...
19) Message boards : Web interfaces : Mauritius - Missing country for account information (Message 2661)
Posted 18 Jan 2006 by Paul D. Buck
Post:
If Janus did it, it is in the baseline and the next time the projects update the site software the change will be on all sites.
20) Message boards : BOINC Manager : Upload problem (Message 2660)
Posted 18 Jan 2006 by Paul D. Buck
Post:
The basic message is described in the Wiki, the error code 400.

I just added more, but, fundamentally, the problem is a networking issue. As I said over in the Rosetta@Home forums, I am not a super networking type, but, problems with firewalls, routers, etc. can all play a part.

There are some networking troubleshooting tips in the Wiki "How-To" to see if you can find some more information about what is happening ... but, what it looks like to me is a truncated session.


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.