Welcome to MilkyWay@home

Posts by Jim Wilkins

21) Message boards : Number crunching : post milkyway_x86_84-apple-darwin problems here (Message 8669)
Posted 19 Jan 2009 by Jim Wilkins
Post:
I certainly did not know that.

Thanks,
Jim
22) Message boards : Number crunching : post milkyway_x86_84-apple-darwin problems here (Message 8663)
Posted 19 Jan 2009 by Jim Wilkins
Post:
I have a 32-bit MacPro and I just got a WU with the 64-bit app. Is that correct?

Jim
23) Message boards : Number crunching : App v0.10 (Message 8661)
Posted 19 Jan 2009 by Jim Wilkins
Post:
I have seen very little increase or decrease from v. 0.7 on either my Mac or PC.

Jim
24) Message boards : Number crunching : Are we doing science yet? (Message 7889)
Posted 20 Dec 2008 by Jim Wilkins
Post:
Impressive!
25) Message boards : Number crunching : Are we doing science yet? (Message 7811)
Posted 17 Dec 2008 by Jim Wilkins
Post:
Yes. Instead of ten hours, I am seeing 30 minutes or less.

Jim
26) Message boards : Number crunching : Are we doing science yet? (Message 7778)
Posted 15 Dec 2008 by Jim Wilkins
Post:
So currently I am seeing a 20x improvement in my crunch times. Is that due to the optimized app, shorter WUs or both?

Thanks,
Jim
27) Message boards : Number crunching : Are we doing science yet? (Message 7711)
Posted 13 Dec 2008 by Jim Wilkins
Post:
I have lost the bubble here. Are we still in a test phase with the optimized apps or are we doing science with them? Just curious.

Jim
28) Message boards : Number crunching : post milkyway_i686_apple darwin problems here (Message 7183)
Posted 1 Dec 2008 by Jim Wilkins
Post:
FYI,

The latest set of WUs (long ones?) are taking about 35 minutes on my quad Mac.
29) Message boards : Number crunching : post milkyway_i686_apple darwin problems here (Message 7122)
Posted 1 Dec 2008 by Jim Wilkins
Post:
Errors fixed in v 0.6.
30) Message boards : Number crunching : post milkyway_i686_apple darwin problems here (Message 6969)
Posted 29 Nov 2008 by Jim Wilkins
Post:
Got errors both reading and writing checkpoint file.
31) Message boards : Number crunching : Running test apps (Message 6950)
Posted 29 Nov 2008 by Jim Wilkins
Post:
Erk! Sounds too hard for this computer challenged guy.

I knew about the response to the accumulated debt and that works fine on my 4 core Mac, but doesn't work at all for my poor single CPU not-very-fast PC. The combination of long runtimes and short deadlines really honks up the other projects on the PC.

Jim

Yes, but you have to make an app_info file to specify just the current MW test app yourself.

You do realize though, that even though the current 'production' app drives the CC into EDF, the accumulated debt that generates will shut down MW DL's for a while after the task completes so your host has a chance to get back to it's assigned Resource Share.

<edit> I also should have mentioned that you have to keep an eye on things, manually download and install the current test app, then modify the app_info file as things develop as well. Almost as annoying as what you are seeing! ;-)

Current BOINC server packages do allow participants to specify what apps to run from the web page UI, but that hasn't been enabled here at MW.

Alinator

32) Message boards : Number crunching : Running test apps (Message 6913)
Posted 28 Nov 2008 by Jim Wilkins
Post:
Is there any way to run JUST the test apps? I am aborting all of the version 1.22 apps as they my machines to go into instant high priority mode.

Thanks,
Jim
33) Message boards : Number crunching : New Mac Intel Performance (Message 3049)
Posted 6 Apr 2008 by Jim Wilkins
Post:
Based one one sample, there appears to be no performance improvement in 1.22 from 1.21. Was there supposed to be?

Thanks,
Jim
34) Message boards : Number crunching : known issues (Message 2438)
Posted 21 Mar 2008 by Jim Wilkins
Post:
The Mac Intel App seems not to checkpoint. With the short running times, is it supposed to?

Thanks.
Jim
35) Message boards : Number crunching : what platforms currently are getting errors? (Message 2105)
Posted 8 Mar 2008 by Jim Wilkins
Post:

I am seeing these user requests also on version 1.21 on Mac OSX Intel. In fact, I have 7 of them in 90 minutes.

Jim


well don't know about the mac stuff but just let it run. As far as i know there are no issues like those on older win platforms that cause computational errors.

Though the mac app seems quite slow...


I have been finishing a task usually in around 750 seconds.

This strange user request glitch has only started since 1.21. Seems to have no adverse effect.

Jim



the user request glitch might be because of putting <next_rpc_delay>600</next_rpc_delay> in the servers config.xml to hopefully let people download WUs faster. it looks like this didn't fix the problem so i'll be removing it.


You know, FWIW, the message says "requested by project", not "requested by user".

Jim

36) Message boards : Number crunching : what platforms currently are getting errors? (Message 2080)
Posted 7 Mar 2008 by Jim Wilkins
Post:

I am seeing these user requests also on version 1.21 on Mac OSX Intel. In fact, I have 7 of them in 90 minutes.

Jim


well don't know about the mac stuff but just let it run. As far as i know there are no issues like those on older win platforms that cause computational errors.

Though the mac app seems quite slow...


I have been finishing a task usually in around 750 seconds.

This strange user request glitch has only started since 1.21. Seems to have no adverse effect.

Jim
37) Message boards : Number crunching : what platforms currently are getting errors? (Message 2064)
Posted 7 Mar 2008 by Jim Wilkins
Post:

Even if I had done that it should surely make only one attempt, not repeatedly try, especially as it is making contact with the server.
But that's the issue, I actually didn't do anything. In fact I was at work and no one could have done it.

In trying to restart the pc I had "redir32.dll" (or "redir.dll" - not sure now) unable to end and had to hard power off. This has made the pc very unhappy and I'm still trying to get it to start up properly!

Safe to say, I'm sure the repeated connecting will be gone once the pc starts up again.

Thanks
Rod


Well that's kinda strange cuz a "user request" can only be triggered by hitting the update button ... have a gost at home ? LOL

However just let that thing run without user/gost intitialised updates ;)

I am seeing this user request messages also without hitting update. Running version 1.21 on OSX Intel.

Jim
There should be a new app out soon that'll fix the issues on W2K,NT4 and Win98...




I am seeing these user requests also on version 1.21 on Mac OSX Intel. In fact, I have 7 of them in 90 minutes.

Jim
38) Message boards : Number crunching : App 1.16 (Message 1889)
Posted 5 Mar 2008 by Jim Wilkins
Post:
Should I abort 32 bit Windows 1.16 tasks?

Jim
39) Message boards : Number crunching : Unit caught in a loop (Message 1238)
Posted 1 Jan 2008 by Jim Wilkins
Post:
As mentioned in other threads, the WU will progress to 100% then the percent complete will reset and count up slowly to about .150%. The first part is the WU crunching and the reset does some kind of internal validation or something. Anyway it is perfectly normal for this to happen.

There are issues right now of WUs not d/l properly and clogging the system.


Sounds good. I have not seen the d/l issues, but then, I only joined today.

Jim
40) Message boards : Number crunching : Unit caught in a loop (Message 1232)
Posted 1 Jan 2008 by Jim Wilkins
Post:
I am seeing a possibly related issue. When a WU executes past it's estimated running time, it resets the time to complete. At this point, the progress indicator increases very little and the time to complete increases slowly. However, within a minute or two (wallclock time) the WU finishes. THe problem seems cosmetic, but I really can't tell.

Jim


Previous 20

©2020 Astroinformatics Group