FahMon (multi-platform app to monitor various F@h clients)

[bollix47]

Tried the latest svn on Ubuntu 8.04 and got the following error when starting FAHMON:

Code: Select all

fahmon: error while loading shared libraries: libwxcurl.so.1: cannot open shared object file: No such file or directory
bollix@challenger:~/downloads/temp/fahmon$ locate libwxcurl.so.1
/home/bollix/downloads/temp/fahmon/wxcurl/src/.libs/libwxcurl.so.1
/home/bollix/downloads/temp/fahmon/wxcurl/src/.libs/libwxcurl.so.1.0.0
/usr/local/lib/libwxcurl.so.1
/usr/local/lib/libwxcurl.so.1.0.0
bollix@challenger:~/downloads/temp/fahmon$ echo $PATH
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games

I did try putting a link in /usr/local/bin to the file in /usr/local/lib but that didn't help the problem.

I also tried both the GnuTLS and the OpenSSL versions of the libcurl - dev from synaptic and that made no difference either.

[uncle_fungus]

You don't actually need to dev headers for libcurl, but they won't hurt anyhow.

My guess is your ldconf doesn't include /usr/local/lib, or you need to run `sudo ldconfig` first to update the cache.

If you run fahmon from the src directory it should work (it uses some trickery to link the libraries at runtime), but the "better" solution would be to run configure with --prefix=/usr to get FahMon to install to /usr instead of /usr/local

There are still a few things I need to iron out in the config scripts now that I'm including additional libraries.

[bollix47]

Using 'sudo ldconfig' helped move things along but after the screen opened up it complained about not being able to determine length of some file. It did ask to send you a debug report which I have done and FAHMON simply closed afterwards.

My older version of svn still 'makes' and works well, so I think I should just wait until you've finished working out the issues you mentioned.

[dempaSD]

Indeed, the SVN code has been handling this very well for a while now, FahMon just reads the last 4KiB of the log and ignores the rest.

Thanks for reply! Any news about when we could see a new official release for Win? I'd rather contribute with something than having to check out code from svn and compile myself right now..

[uncle_fungus]

A new release is due very soon, there are a few things to finalise, but this should clear up all (hopefully) of the outstanding bugs.

[codysluder]

I have not attempted to read your code, so I'm not sure how you handle the names for UNKNOWN cores, but why is it imbeddded in the code? Couldn't you provide an external file that could be updated either by just downloading that file or with a text editor. Then we wouldn't need to wait for you to compile/post/distribute a new version.

[uncle_fungus]

I'm not sure how you handle the names for UNKNOWN cores, but why is it imbeddded in the code?

Purely historical reasons.

Couldn't you provide an external file that could be updated either by just downloading that file or with a text editor. Then we wouldn't need to wait for you to compile/post/distribute a new version.

Potentially, yes, but not for this release.

[SantaFe]

Howdy !!

Do have a couple of questions. I went & compiled the source code for the Linux version (After making sure I had ALL the necessary components first!) and followed the instructions (./configure, then make & finally sudo make install). Had no problems at all compiling, no error messages, but when I click on the icon to start it, a couple of boxes pop up saying:

ASSERT INFO:
../src/generic/hyperlink.cpp(80): assert "!url.empty() || !label.empty()" failed in Create(): Both URL and label are empty ?

And when I click on backtrace I get:

BACKTRACE:
[1] MainDialog::CreateLayout()
[2] MainDialog() /home/craig/My-Downloads/FahMon-2.3.4/src/mainDialog.cpp:178
[3] MainDialog::CreateInstance() /home/craig/My-Downloads/FahMon-2.3.4/src/mainDialog.cpp:219

It then runs like it should except when F@H client gets a new project when the previous one finishes, FAHMON closes (no error message) & I have to restart it.

What did I do? did I forgot something? I don't mind the pop up boxes so much, I just click continue & go on, but the part where it quits when F@H uploads the WU & gets a new one is kinda weird.

[uncle_fungus]

Howdy !!

Do have a couple of questions. I went & compiled the source code for the Linux version (After making sure I had ALL the necessary components first!) and followed the instructions (./configure, then make & finally sudo make install). Had no problems at all compiling, no error messages, but when I click on the icon to start it, a couple of boxes pop up saying:

ASSERT INFO:
../src/generic/hyperlink.cpp(80): assert "!url.empty() || !label.empty()" failed in Create(): Both URL and label are empty ?

And when I click on backtrace I get:

BACKTRACE:
[1] MainDialog::CreateLayout()
[2] MainDialog() /home/craig/My-Downloads/FahMon-2.3.4/src/mainDialog.cpp:178
[3] MainDialog::CreateInstance() /home/craig/My-Downloads/FahMon-2.3.4/src/mainDialog.cpp:219

That means you're using a debug build of wxGTK. The warning is nothing to worry about, its just complaining that the username and team number URLs are set to be blank when the program intialises.

It then runs like it should except when F@H client gets a new project when the previous one finishes, FAHMON closes (no error message) & I have to restart it.

What did I do? did I forgot something? I don't mind the pop up boxes so much, I just click continue & go on, but the part where it quits when F@H uploads the WU & gets a new one is kinda weird.

That definately shouldn't be happening. If you can run fahmon inside of gdb that would help greatly as it will tell me where in the code the crash is occuring.

Code: Select all

gdb src/fahmon <enter>
run <enter>

[bollix47]

I just tried the svn on another computer to see if anything was different from the original computer I tried it on and fahmon is working now with one exception.

All the GPU clients show zero PPD and the ETA on all of them is the current time.

The SMP clients all show the correct ETA and PPD.

Hope this information is helpful.

[uncle_fungus]

I just tried the svn on a another computer to see if anything was different from the original computer I tried it on and fahmon is working now with one exception.

All the GPU clients show zero PPD and the ETA on all of them is the current time.

The SMP clients all show the correct ETA and PPD.

Hope this information is helpful.

Are the GPU clients running core_14 WUs? (Is the database up to date)

PS. I'm glad more people are finally trying the SVN builds. Feedback is really important so I can fix the bugs ready for release.

[bollix47]

Two of the GPU clients are core 14 and the other two are not.

Yes, the database is up-to-date and all the GPU clients show the correct point values of 420, 353, 511.

The progress percentages are updating correctly.

It's just the ETA and PPD that are wrong. If I reload any of the 4 GPU clients the ETA changes to the current time and the PPD stays at zero.

I'm including the messages window and even though it points to an error reading the log files for the SMP clients they are all showing correct information in the monitoring window.

Code: Select all

[04/03/09 - 10:58:51.784]   FahMon 2.3.4 is licenced under the GNU GPL v2
[04/03/09 - 10:58:51.784]   Copyright (C) 2003-2007 François Ingelrest
[04/03/09 - 10:58:51.784]   Copyright (C) 2007-2009 Andrew Schofield
[04/03/09 - 10:58:51.784]   
[04/03/09 - 10:58:51.784]   FahMon contains code from qd, also licenced under the GNU GPL v2
[04/03/09 - 10:58:51.784]   Copyright (C) 2002-2005 Richard P. Howell IV.
[04/03/09 - 10:58:51.784]   Copyright (C) 2005-2009 Sebastiaan Couwenberg
[04/03/09 - 10:58:51.784]   
[04/03/09 - 10:58:51.872] X Error while reading /home/bollix/.gvfs/fah on georgina/smp/work/logfile_05.txt
[04/03/09 - 10:58:52.086] X Error while reading /home/bollix/.gvfs/fah on endeavour/smp/work/logfile_04.txt
[04/03/09 - 10:58:52.189] X Error while reading /home/bollix/.gvfs/bollix_on_challenger on ubuntu_server/fah/smp/work/logfile_01.txt
[04/03/09 - 11:03:51.935] X Error while reading /home/bollix/.gvfs/fah on georgina/smp/work/logfile_05.txt
[04/03/09 - 11:03:52.161] X Error while reading /home/bollix/.gvfs/fah on endeavour/smp/work/logfile_04.txt
[04/03/09 - 11:03:52.259] X Error while reading /home/bollix/.gvfs/bollix_on_challenger on ubuntu_server/fah/smp/work/logfile_01.txt

btw the speed of loading the clients is now very fast and the calculation predictions for the SMP WUs with odd sizes like 250001 and 249999 are now correct.

[Eveofwar]

I absolutely love this application, thank you to the developer and the team that keeps it up and running ! Keep up the good work guys

[uncle_fungus]

Two of the GPU clients are core 14 and the other two are not.

Yes, the database is up-to-date and all the GPU clients show the correct point values of 420, 353, 511.

The progress percentages are updating correctly.

It's just the ETA and PPD that are wrong. If I reload any of the 4 GPU clients the ETA changes to the current time and the PPD stays at zero.

If you could email me a copy of the log from one of the broken clients I can't probably figure out whats happening (andrew_s {at} fahmon {dot} net).

I'm including the messages window and even though it points to an error reading the log files for the SMP clients they are all showing correct information in the monitoring window.

Code: Select all

[04/03/09 - 10:58:51.784]   FahMon 2.3.4 is licenced under the GNU GPL v2
[04/03/09 - 10:58:51.784]   Copyright (C) 2003-2007 François Ingelrest
[04/03/09 - 10:58:51.784]   Copyright (C) 2007-2009 Andrew Schofield
[04/03/09 - 10:58:51.784]   
[04/03/09 - 10:58:51.784]   FahMon contains code from qd, also licenced under the GNU GPL v2
[04/03/09 - 10:58:51.784]   Copyright (C) 2002-2005 Richard P. Howell IV.
[04/03/09 - 10:58:51.784]   Copyright (C) 2005-2009 Sebastiaan Couwenberg
[04/03/09 - 10:58:51.784]   
[04/03/09 - 10:58:51.872] X Error while reading /home/bollix/.gvfs/fah on georgina/smp/work/logfile_05.txt
[04/03/09 - 10:58:52.086] X Error while reading /home/bollix/.gvfs/fah on endeavour/smp/work/logfile_04.txt
[04/03/09 - 10:58:52.189] X Error while reading /home/bollix/.gvfs/bollix_on_challenger on ubuntu_server/fah/smp/work/logfile_01.txt
[04/03/09 - 11:03:51.935] X Error while reading /home/bollix/.gvfs/fah on georgina/smp/work/logfile_05.txt
[04/03/09 - 11:03:52.161] X Error while reading /home/bollix/.gvfs/fah on endeavour/smp/work/logfile_04.txt
[04/03/09 - 11:03:52.259] X Error while reading /home/bollix/.gvfs/bollix_on_challenger on ubuntu_server/fah/smp/work/logfile_01.txt

They are non-fatal errors, however it does pose a slight problem in that the client creates the logfile_xx with permissions that deny you read access if the client isn't started under your username. I may have to re-think the parsing of that file to access the core version.

btw the speed of loading the clients is now very fast and the calculation predictions for the SMP WUs with odd sizes like 250001 and 249999 are now correct.

Excellent.

I absolutely love this application, thank you to the developer and the team that keeps it up and running ! Keep up the good work guys

Thanks for the support!

[SantaFe]

That definately shouldn't be happening. If you can run fahmon inside of gdb that would help greatly as it will tell me where in the code the crash is occuring.

Code: Select all

gdb src/fahmon <enter>
run <enter>

Yep....using Ubuntu 9.04 (jaunty)
GNOME: 2.25.92 (Ubuntu 2009-03-03)
Kernal: 2.6.28-8-generic (#26-Ubuntu SMP Wed Feb 25 04:28:54 UTC 2009)

Will find out Tomorrow, that's when the next WU is expected to be sent. Now if only I can remember NOT to close that terminal window! Thanks!

Actually, even the second problem isn't really that big, I just reclick the FahMon Icon & it's back, but just wanted to know WHY!