Folding Forum

Thanks Uncle!
Great work!

uncle_fungus:

After updating to the 2.3.3 release version, I've noticed that the client updates "hang" intermittently, e.g. they stop auto-updating (I'm using the default autoupdate every 3 minutes). There's no error message or anything, the client list just looks normal but no updates occur. Nothing happens if I reload a specific client (F5) or reload all (F6). The only way to get going again is to close FahMon and restart it. Then it picks up where it's supposed to be.

Is this something you've seen before? I'm running on Vista Ultimate 64-bit (up-to-date), and I used the zipped version to unpack in my old 2.3.2b folder (which never displayed this problem).

Please let me know if there's anything you want me to try or post, thanks!

2.3.3 seems to be functioning perfectly.

One *very* tiny suggestion. During initial setup when adding a new client the first time you open the Client Information pop-up box the cursor is flashing in the Name box where it should be. When adding a second or later new client the cursor is gone. Those of us (like me) who are not always paying attention as closely as we should may type in the entire name of the second client before realizing the cursor is not in the box this time and is therefore not accepting the typing of the new client name.

Thanks for all of your efforts providing us with your very useful tool.

Here's an addendum to my post above, from the FahMon messages window. As you can see things start to get wrong, and eventually only one SMP client is updated and none of the others. No errors displayed in the Fah window, everything looks dandy.
At the 03:34:08 mark I pressed F6 a couple of times, and you'll see it only reloads one client (out of 7).
These are 3 machines on a local gigabit network, all shares working fine. All are running Vista Ultimate 64-bit.

It appears to me that the problem starts here:
[02/11/08 - 03:05:19] Possible clock asynchrony detected!

After that, the next update starts dropping clients, and the update after that only one client is alive.

osgorth wrote:After updating to the 2.3.3 release version, I've noticed that the client updates "hang" intermittently, e.g. they stop auto-updating (I'm using the default autoupdate every 3 minutes). There's no error message or anything, the client list just looks normal but no updates occur. Nothing happens if I reload a specific client (F5) or reload all (F6). The only way to get going again is to close FahMon and restart it. Then it picks up where it's supposed to be.

Mine hangs and stays hanging, even if I restart it, reload it, delete it and restart it. I am using Vista 32-bit.

Edit: Here is part of the FahMon log:

[01/11/08 - 23:16:25] Reloading GPU One
[01/11/08 - 23:16:25] Possible clock asynchrony detected!
[01/11/08 - 23:16:25] ! GPU One seems to have hung : Elapsed time is 1380m and limit is 90m
[01/11/08 - 23:16:25] GPU One is on frame 30
[01/11/08 - 23:16:25] GPU One [ETA] | avg=00:01:19 | ref=00:01:29 | adj=00:01:29 | left=01:42:21
[01/11/08 - 23:16:29] Reloading GPU Two
[01/11/08 - 23:16:29] Possible clock asynchrony detected!
[01/11/08 - 23:16:29] ! GPU Two seems to have hung : Elapsed time is 1381m and limit is 90m
[01/11/08 - 23:16:29] GPU Two is on frame 16
[01/11/08 - 23:16:29] GPU Two [ETA] | avg=00:03:21 | ref=00:03:25 | adj=00:03:25 |

HI,

For those clients that hang and get "Possible clock asynchrony detected!" goto preferences and check the box for ignore asynchronous clocks and that will fix the problem.

L8R

I just installed 2.3.3 and it looks great. No problems here.
I would like to suggest adding an F key toggle for Show/Hide Info Panel. You have F8 & F9 for the other toggles. It'd save a few mouse clicks.

Thanks,,, Dave

tbk-aracthebold wrote:HI,

For those clients that hang and get "Possible clock asynchrony detected!" goto preferences and check the box for ignore asynchronous clocks and that will fix the problem.

L8R

That option is already set. Sorry, I forgot to mention that!

The older 2.3.2beta2 is showing the same asynchronous clock "blue box" for the folding clients too. As we are in the process of the world-wide switching of the time clocks back to Standard Time I suspect that might be the cause of the problem.

For people getting hung clients, make sure that the timezone is set correctly in FahMon's preferences, and if your various machine's clocks are out of sync, turn on "ignore asynchronous clocks".

osgorth wrote:Here's an addendum to my post above, from the FahMon messages window. As you can see things start to get wrong, and eventually only one SMP client is updated and none of the others. No errors displayed in the Fah window, everything looks dandy.
At the 03:34:08 mark I pressed F6 a couple of times, and you'll see it only reloads one client (out of 7).
These are 3 machines on a local gigabit network, all shares working fine. All are running Vista Ultimate 64-bit.

It appears to me that the problem starts here:
[02/11/08 - 03:05:19] Possible clock asynchrony detected!

After that, the next update starts dropping clients, and the update after that only one client is alive.

Code: Select all

[02/11/08 - 03:02:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:02:19]   Reloading Core 2 Quad Extreme 3.81GHz - smp
[02/11/08 - 03:02:19]   Sapphire Radeon 3870 - gpu is on frame 22
[02/11/08 - 03:02:19]   Sapphire Radeon 3870 - gpu [ETA] | avg=00:03:39 | ref=00:03:46 | adj=00:02:57 | left=04:50:51
[02/11/08 - 03:02:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:02:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:37:10 | left=88:52:46
[02/11/08 - 03:02:19]   Core 2 Quad Extreme 3.81GHz - smp is on frame 9
[02/11/08 - 03:02:19]   Core 2 Quad Extreme 3.81GHz - smp [ETA] | avg=00:19:38 | ref=00:23:09 | adj=00:11:25 | left=34:55:14
[02/11/08 - 03:02:19]   Reloading Gigabyte 8800GTS (G92) - gpu
[02/11/08 - 03:02:19]   Reloading Core 2 Quad Extreme 3.81GHz - cpu
[02/11/08 - 03:02:19]   Gigabyte 8800GTS (G92) - gpu is on frame 65
[02/11/08 - 03:02:19]   Gigabyte 8800GTS (G92) - gpu [ETA] | avg=00:01:17 | ref=00:01:09 | adj=00:00:23 | left=00:39:52
[02/11/08 - 03:02:19]   Core 2 Quad Extreme 3.81GHz - cpu is on frame 87
[02/11/08 - 03:02:19]   Core 2 Quad Extreme 3.81GHz - cpu [ETA] | avg=00:09:25 | ref=00:09:29 | adj=00:01:05 | left=02:02:12
[02/11/08 - 03:05:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:05:19]   Reloading Sapphire Radeon 4870x2 - gpu
[02/11/08 - 03:05:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:05:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:40:10 | left=88:49:46
[02/11/08 - 03:05:19]   Possible clock asynchrony detected!
[02/11/08 - 03:05:19]   Sapphire Radeon 4870x2 - gpu has been marked as having a asynchronous clock.
[02/11/08 - 03:05:19]   Sapphire Radeon 4870x2 - gpu is on frame 81
[02/11/08 - 03:05:19]   Sapphire Radeon 4870x2 - gpu [ETA] | avg=00:02:12 | ref=00:02:14 | adj=00:02:14 | left=00:40:12
[02/11/08 - 03:08:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:08:19]   Reloading Gigabyte 8800GTS (G92) - gpu
[02/11/08 - 03:08:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:08:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:43:10 | left=88:46:46
[02/11/08 - 03:08:19]   Gigabyte 8800GTS (G92) - gpu is on frame 70
[02/11/08 - 03:08:19]   Gigabyte 8800GTS (G92) - gpu [ETA] | avg=00:01:17 | ref=00:01:09 | adj=00:00:36 | left=00:33:54
[02/11/08 - 03:11:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:11:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:11:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:46:10 | left=88:43:46
[02/11/08 - 03:14:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:14:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:14:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:49:10 | left=88:40:46
[02/11/08 - 03:17:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:17:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:17:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:52:10 | left=88:37:46
[02/11/08 - 03:20:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:20:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:20:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:55:10 | left=88:34:46
[02/11/08 - 03:23:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:23:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:23:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=00:58:10 | left=88:31:46
[02/11/08 - 03:26:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:26:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:26:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:01:10 | left=88:28:46
[02/11/08 - 03:29:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:29:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:29:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:04:10 | left=88:25:46
[02/11/08 - 03:32:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:32:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:32:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:07:10 | left=88:22:46
[02/11/08 - 03:34:08]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:34:08]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:34:08]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:08:59 | left=88:20:57
[02/11/08 - 03:34:11]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:34:11]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:34:11]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:09:02 | left=88:20:54
[02/11/08 - 03:35:19]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:35:19]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:35:19]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:10:10 | left=88:19:46
[02/11/08 - 03:35:28]   Reloading Pentium D 3.2GHz - smp
[02/11/08 - 03:35:28]   Pentium D 3.2GHz - smp is on frame 26
[02/11/08 - 03:35:28]   Pentium D 3.2GHz - smp [ETA] | avg=00:48:30 | ref=01:12:34 | adj=01:10:19 | left=88:19:37

Hmm, that doesn't look good. Thanks for reporting it, it looks to me like the code to prevent multiple refreshes of the same client in quick succession is removing clients from the update queue. I'll see what I can do to fix it.

Thanks Uncle!
Just to clarify: this happens even if I'm not reloading manually. If I leave FahMon for a few hours, it's usually hung up like this when I get back. So it's safe to say it's not triggered by pressing F5/F6.

Just wanted to confirm that the reload 'problem' is happening here as well. After using the F5/F6 keys a few times the updating stops.

I thought there was something wrong as every now and then I need to restart fahmon 2.3.3

Here the same happens as with osgorth and bollix47. However, to narrow the problem down, I should tell that I notice it for my SMP- and GPU2-clients running locally on the same box where FahMon runs.
So I guess TZ-/daylight savings-issues can be ruled out for my setup.

Some details:
- FahMon 2.3.3
- SMP client - v.6.23 r1 MPICH
- GPU2 - v.6.20 r1
- Windows Server 2003 R2 SP2
- Intel Q6600
- 4 GB RAM
- Nvidia 9800GT

Corné