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

[uncle_fungus]

Hi, I translated 2.3.1 to Spanish. Is there need for an updating ? Let me know.

best regards,

Trotador

There are a few minor additions in 2.3.2, if you could email the 2.3.1 translation to me that would be great. Thanks.

[billclo]

I've been seeing sporatic issue with Fahmon 2.32, wherein it starts taking up an entire CPU core all for itself. Memory consumption seems to be about 8MB, but soaking up an entire core is weird. Once it was when a client computer was down and I hadn't removed it from the list to be monitored. Lately, it's been just whenever.

My main machine is slow enough (A64 X2 3800) on recent proteins, but Fahmon soaking an entire core is plain unacceptable. Any ideas guys?

[Gerby19]

Hi billclo!

The problem could be the experimental reload system (FahMon > Preferences > Monitoring). If it is activated, try to run FahMon without it.

Greetings
Gerby

[pfb]

uncle_fungus,

Any update on Ticket 104 (FahMon refuses to restore)? Got this bug for the past few days now

/edit - managed to get the window showing again - seems it shows as 1x1 window when the bug hits (I have dual monitors and use Ultramon - sending FahMon to the 2nd window managed to show the title bar of the window and manage to resize from there)

1 thing I did notice is that when this bug hits, the web output doesn't work either - all the clients show as inaccessible...

[uncle_fungus]

I haven't had the time to work on FahMon the past few days. In order to fix this, I think the whole position/size saving system needs to be re-written from scratch to cope with the differences in the way the 3 OSes handle things, as what I've hacked together so far obviously doesn't cut it (on Win32).

The first time fahmon creates the web output will almost always generate a list of inaccessible clients. The next time it updates the web output will be correct. This is completely independent of what state the window is in.

[uncle_fungus]

Linked to from here: viewtopic.php?f=10&t=1357&p=11816#p11816 to keep threads on topic.

I think it may depend on how FahMon gets the data that is displayed in the right-hand panel - - but this should be discussed in the 3rd party forum.

The username and team data is grabbed from the queue, so I sincerely hope that beta5 hasn't altered the format. Try running qd on the queue and see if it can decipher the info.

Is FahMon capable of dealing with each client being assigned to a different team or a different UserName?

Yes it is.

[muziqaz]

About that window thingy. I had it several times, but I found out which files is responsible for this.
When I 1st installed fahmon, I backed up all config files, that is: benchmarks.dat, clientsab.txt, prefs.dat, projects.dat. And when that bug appeared I deleted config dir from program directory, and put back files from back up. And everything worked again, i could see main window. Next time when it happened I replaced each file from config dir one by one. The one that gave me back visible window was prefs.dat. If you do not have backup, you just delete that file, and launch fahmon again, and reconfigure the program.

i hope it will help in fixing this

[uncle_fungus]

The one that gave me back visible window was prefs.dat. If you do not have backup, you just delete that file, and launch fahmon again, and reconfigure the program.

i hope it will help in fixing this

Yeah, that works because deleting prefs.dat deletes the saved window states which forces FahMon to use the hard-coded default values, which work.

The problem is, that for some reason, the minimised/maximised states are not being handled correctly, which causes FahMon to start with its main window in a stupid state.

[Adanorm]

Hello, i've got 2 FAHMON on 2 server, dans the two say me at the upgrade of the proteine list that Psummary is not correctly formated !

Is it a FAHMON issue or stanford ?

[uncle_fungus]

Hello, i've got 2 FAHMON on 2 server, dans the two say me at the upgrade of the proteine list that Psummary is not correctly formated !

Is it a FAHMON issue or stanford ?

It happens whenever there are missing data fields on psummary. It should continue processing the rest of the data though.

[uncle fuzzy]

I've had a minor annoyance on my quad. It's the only one running 2.3.1. It ramdomly lists a client user name as anonymous (0) instead of uncle_fuzzy (52286). It doesn't reconfigure the client, so it's just a cosmetic problem. Closing and restarting usually fixes it. Would changing to a newer version take care of this?

edit- q6600 XP Home

The rest of my machines are all running 2.3.0 because I prefer to have ETA in D/H/M, but I want the download and deadlines in month/day. I haven't been able to make the newer ones do this. Changing the prefs changes all of them.

[Foxery]

Minor bug - I don't see this mentioned here or on the Trac. I'm running v2.3.2.

Time Remaining calcs for WUs whose frames are not listed in 1% increments exhibit strange behavior. Specifically, I've had projects 3905-3908 a few times lately. The percentage shown in the list of clients is correct, and the PPD calculation looks correct, but the estimated time remaining is always many hours more than reality. It seems to either remain roughly the same all day long, or even increase slightly.

Beyond that, I love this tool for watching my office borg. Keep up the good work!

[uncle_fungus]

Minor bug - I don't see this mentioned here or on the Trac. I'm running v2.3.2.

Time Remaining calcs for WUs whose frames are not listed in 1% increments exhibit strange behavior. Specifically, I've had projects 3905-3908 a few times lately. The percentage shown in the list of clients is correct, and the PPD calculation looks correct, but the estimated time remaining is always many hours more than reality. It seems to either remain roughly the same all day long, or even increase slightly.

Beyond that, I love this tool for watching my office borg. Keep up the good work!

Interesting. It should work correctly, but if not I'll take a look and see if I can work out what's going on.
I'm pretty busy in RL at the moment which is why there hasn't been much (any) development activity recently.

[ArVee]

I apologize if this has already been covered; I did search Help and the forum before posting. FahMon is reporting a Linux SMP unit that is folding under VMWare on a Windows machine as "hung." The unit is not hung, and FahMon even shows the updated percentage progress while showing the ETA as Hung. Any ideas?

[bollix47]

I've had this happen a few times too. What I found was that the time on the machine I was monitoring was a few seconds faster than the computer doing the monitoring.

Usually I just wait until the FAHMON computer goes to the next minute and then right-click on the 'hung' entry and select reload.

I believe what's happening is that because the time on the machine being monitored is in the future FAHMON thinks it's hung when it's not.