Help! Strange network behavior using WideFS and PM

[turboarrow]

Hello

Recently, I bought two last generation Compaq Desktop PCs in order to use them as clients within my projectmagenta-Network. On the third PC, I'm having a 3Ghz PC running FSX and wideserver (newest versions of server and client for FSX) only.

At the beginning, I had big problems: usind MCP and CDU on client 1 and the Boeing PFD on client 2, a modification in a flightplan was visible 10-30 seconds later on the PFD at the other client. Generally, the whole stuff was stuttering quite heavily.

So I visited this forum and tried to retrace all tips and hints I've found for similar problems:

- deactivation of IEEE802

- deactivation of index services

- static IP addresses for server and clients

- Adaption of Norton Internet security 2007

Result: it became really better, but the problem is still not 100% solved.

When I start my system, everything works fine at the beginning. Immediately after departure, the PFD start stuttering and after at least the fifth flightpan-update (e.g. direct-to functions), the PFD stops updating information and abends as well as the CDU und MCP.

I'm sorry but I'm not able to locate the problem anywhere else. Is it maybe the Hub (will be replaces by a gigabyte-switch soon)? Or any other network setting?

Please find below the logs of the Wideserver, MCP and PFD:

WideServer:

********* WideServer.DLL Log [version 7.09] *********

Blocksize guide = 4096 (double allowed)

Date (dmy): 15/04/07, Time 13:28:26.156: Server name is MEDIONPC

15250 Initialising TCP/IP server

15250 Initialising IPX/SPX server

15266 IPX/SPX socket() failed [Error=10047] Address family not supported by protocol family

15266 Failed to start IPX/SPX Server

15266 Initialising UDP/IP server

16141 Broadcasting service every 1000 mSecs

16672 Incoming connection Accepted ok (skt=4184) TCP

16672 Incoming connection Accepted ok (skt=4196) TCP

16844 Connected to computer "CLIENT2" running WideClient version 6.710 (skt=4184) TCP

16938 Connected to computer "CLIENT1" running WideClient version 6.710 (skt=4196) TCP

2659266 Error 10053: client socket disconnected at Client: removing (skt=4184) TCP

2782907 Error 10054: client socket disconnected at Client: removing (skt=4196) TCP

2801391 Close signalled to clients

2802500 Closing down now ...

Memory managed: Offset records: 2018 alloc, 2017 free

Read buffer usage: 24408 alloc, 24408 free, max in session: 1

Write buffer usage: 101192 alloc, 101192 free, max in session: 1

Throughput maximum achieved: 58 frames/sec, 21913 bytes/sec

Throughput average achieved for complete session: 17 frames/sec, 4968 bytes/sec

Average receive rate from "CLIENT2": 8 frames/sec, 1233 bytes/sec

Average receive rate from "CLIENT1": 0 frames/sec, 18 bytes/sec

********* Log file closed *********

MCP:

15.04.2007 13:34:38 Program Start #################################################

15.04.2007 13:34:38 MCP Build 418

15.04.2007 13:34:38 EXE Date 08.03.2007 21:04:23

15.04.2007 13:34:44 cautionthroughpass='off'

15.04.2007 13:34:48 getting cdu data from \\CLIENT2\netdir\

15.04.2007 13:34:48 read cdu data from \\CLIENT2\netdir\

...

15.04.2007 14:02:06 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:02:06 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:02:41 Read Error NoLat

15.04.2007 14:03:03 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:03:03 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:04:15 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:04:16 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:05:01 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:05:01 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:05:59 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:05:59 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:06:57 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:06:57 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:07:32 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:07:32 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:07:46 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:07:46 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:09:55 getting cdu data from \\CLIENT2\netdir\

15.04.2007 14:09:55 read cdu data from \\CLIENT2\netdir\

15.04.2007 14:12:35 Program Ended ########################################################

PFD:

15.04.2007 13:27:14 Program Start #############################################

15.04.2007 13:27:14 date:15.04.2007

15.04.2007 13:27:14 Init OK

15.04.2007 13:27:14 Identified

15.04.2007 13:27:14 Loading Main

15.04.2007 13:27:14 Main T-L-H

15.04.2007 13:27:14 Main SetSize

15.04.2007 13:27:14 Init Vars

15.04.2007 13:27:14 Main Loaded

15.04.2007 13:27:14 Main SetSize

15.04.2007 13:27:14 Load Done

15.04.2007 13:27:17 setting up OpenGL

15.04.2007 13:27:17 font installed

15.04.2007 13:27:17 TypeFile b737.txt

15.04.2007 13:27:17 Loading: type.txt

15.04.2007 13:27:17 TypeFile Stamp 18.01.2001 11:57:42

15.04.2007 13:27:17 CDU Not Detected

15.04.2007 13:27:17 copied type.dat file

15.04.2007 13:27:17 v1 vr v2 vref 0 0 0 0 (craft)

15.04.2007 13:27:17 v1 vr v2 vref 112 113 120 117

15.04.2007 13:27:17 Main SetSize

15.04.2007 13:27:17 Main SetSize

15.04.2007 13:27:17 TypeFile type.txt

15.04.2007 13:27:17 Loading: type.txt

15.04.2007 13:27:17 TypeFile Stamp 18.01.2001 11:57:42

15.04.2007 13:27:17 CDU Not Detected

15.04.2007 13:27:17 v1 vr v2 vref 0 0 0 0 (craft)

15.04.2007 13:27:17 v1 vr v2 vref 112 113 120 117

15.04.2007 13:29:06 TypeFile type.txt

15.04.2007 13:29:06 Loading: type.txt

15.04.2007 13:29:06 TypeFile Stamp 18.01.2001 11:57:42

15.04.2007 13:29:06 CDU Not Detected

15.04.2007 13:29:06 v1 vr v2 vref 0 0 0 0 (craft)

15.04.2007 13:29:06 v1 vr v2 vref 112 113 120 117

15.04.2007 13:29:06 FS Path: \\MEDIONPC\fs\

15.04.2007 13:29:06 TypeFile type.txt

15.04.2007 13:29:06 Loading: type.txt

15.04.2007 13:29:06 TypeFile Stamp 18.01.2001 11:57:42

15.04.2007 13:29:06 CDU Not Detected

15.04.2007 13:29:06 v1 vr v2 vref 0 0 0 0 (craft)

15.04.2007 13:29:06 v1 vr v2 vref 134 135 139 133

15.04.2007 13:29:11 Loading Flightplan

15.04.2007 13:29:11 Flightplan Source=2 [0=CDU 1=SBP 2=FS]

15.04.2007 13:29:11 File used >\\MEDIONPC\fsplans\IFR Zurich to Reina Sofia.PLN<

15.04.2007 13:34:44 CDU Instructs Load Flightplan

15.04.2007 13:34:44 Loading Flightplan

15.04.2007 13:34:44 CDU Detected

15.04.2007 13:34:44 Flightplan Source=2 [0=CDU 1=SBP 2=FS]

15.04.2007 13:34:44 File used >\\MEDIONPC\fsplans\IFR Zurich to Reina Sofia.PLN<

15.04.2007 13:34:44 Loading Flightplan

15.04.2007 13:34:44 CDU Detected

...

15.04.2007 14:07:51 CDU Detected

15.04.2007 14:07:52 v1 vr v2 vref 134 135 139 133 (craft)

15.04.2007 14:07:52 v1 vr v2 vref 134 135 139 133

15.04.2007 14:07:54 Loading Flightplan

15.04.2007 14:07:54 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

15.04.2007 14:07:54 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

15.04.2007 14:07:56 Loading Flightplan

15.04.2007 14:07:56 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

15.04.2007 14:07:56 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

15.04.2007 14:08:01 Loading Flightplan

15.04.2007 14:08:01 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

15.04.2007 14:08:02 time to access file 282

15.04.2007 14:10:24 \\CLIENT2\netdir\tmp.pln to ptmp.pln (Copy Error)

15.04.2007 14:11:33 Loading Flightplan

15.04.2007 14:11:33 Flightplan Source=0 [0=CDU 1=SBP 2=FS]

CDU:

15.04.2007 13:33:20 program start: C:\PM\CDU

15.04.2007 13:33:20 local path: C:\PM\CDU

15.04.2007 13:33:20 build: 388

15.04.2007 13:33:22 loading type file

15.04.2007 13:33:22 file found CduType.bin

15.04.2007 13:33:22 file found cdutype.txt

15.04.2007 13:33:23 loading type file done

15.04.2007 13:33:23 file found cdu.ini

15.04.2007 13:33:23 netdir read in INI:\\client2\netdir\

15.04.2007 13:33:23 file found \\client2\netdir\test.cdu

15.04.2007 13:33:23 netdir check ok:\\client2\netdir\

Do you have any idea what could cause this problem?

Thanks a lot in advance and kind regards.

Stephan

[Pete Dowson]

At the beginning, I had big problems: usind MCP and CDU on client 1 and the Boeing PFD on client 2, a modification in a flightplan was visible 10-30 seconds later on the PFD at the other client. Generally, the whole stuff was stuttering quite heavily.

Are you sure these are Network problems and not problems with the graphics drivers for the PFDs?

Immediately after departure, the PFD start stuttering and after at least the fifth flightpan-update (e.g. direct-to functions), the PFD stops updating information and abends as well as the CDU und MCP.

"Abends" meaning what, precisely? Is there an error report, a hang, or what?

I'm sorry but I'm not able to locate the problem anywhere else. Is it maybe the Hub (will be replaces by a gigabyte-switch soon)? Or any other network setting?

No problems according to the WideServer Log, but the other logs you provide are for PM, not I. I'm afraid you need PM support for those. Are there any errors listed in the Wideclient logs on the clients?

And please also check the FSUIPC4.LOG, as any problems with SimConnect may cause stuttering -- not crashes nor hangs however. If your PFDs are crashing the most likely suspect is video drivers on the clients. PM support may be able to help.

BTW, it probably isn't terribly important, but I think it best if you run the latest versions of WideClient on the two client PCs -- 6.72.

Regards

Pete

[Pete Dowson]

One other observation:

... usind MCP and CDU on client 1 and the Boeing PFD on client 2

...

Average receive rate from "CLIENT2": 8 frames/sec, 1233 bytes/sec

Average receive rate from "CLIENT1": 0 frames/sec, 18 bytes/sec

Have you perhaps got Client1 and 2 named the other way around from your description? I would expect next to zero data reception at the server from the PFD program, which is a display only, but qute a bit from the CDU and, especially, the MCP which is controlling the plane during autopilot modes.

Regards

Pete

[turboarrow]

Hello Pete

Thank a lot for your first diagnosis. In the meantime, I’ve updated all drivers (based on the suggestions of driver genius professional) and tried out the whole thing once more – but the system behaviour was as before.

The PFD/ND (yes, you’re right, it’s on client 1, CDU/MCP are on client 2) works OK, bit I’ve remarked that when I e.g. delete a waypoint, the PFD is updated normally, but the ND with the deleted waypoint updates only with a big (up to 30 seconds) time lag.

Sometimes, I already have troubles during the pre-flight-phase (entering the flightplan in the CDU), sometimes, the system reacts badly after departure. The outcome is always the same: the hourglass of the CDU on client 2 as well as of the PFD/ND on client 1 remains and I even have troubles to close the programs (I need to close them for approx. 10 times; I meant that behaviour with “abend”).

The system resources are at no time at the limit, so it shouldn’t be a problem of RAM or CPU.

To the logs:

- Wideclient logs: as I can see, there are no problems reported in the log (pls see attachment)

- Wideserver-log: there are two errors: 10053 and 10054; no idea why this happens.

- CDU/MCP: there are no problems reported in the logs

- PFD: at the end of the log, the CDU was not detected on the other client.

- WideCLient version: I hust realised that I use 6.71 instead of 6.72 – but as you wrote this shouldn’t cause the problems but I will install it before the next test.

- FSUIPC4.LOG: No problem detected, pls see attachment for further details.

Do you still thnk that this is a PM issue (PM told me that it could be a WideFS-problem)?

Thanks a lot in advance and kind regards

Stephan

Logs Test 20070416 1700h.zip

[Pete Dowson]

when I e.g. delete a waypoint, the PFD is updated normally, but the ND with the deleted waypoint updates only with a big (up to 30 seconds) time lag.

What in the PFD is there from your waypoint to be updated? Surely both the PFD display and the ND are in one window, from one program (PFD.EXE) on your client?

Sometimes, I already have troubles during the pre-flight-phase (entering the flightplan in the CDU), sometimes, the system reacts badly after departure. The outcome is always the same: the hourglass of the CDU on client 2 as well as of the PFD/ND on client 1 remains and I even have troubles to close the programs (I need to close them for approx. 10 times; I meant that behaviour with “abend”).

That sounds absolutely awful. What does PM support say?

Have you checked the performance of the PFD display on that client? If you press 'F' (I think) it will display the assorted frame rates - graphic, data and FS.

A process of elimination would be useful too. Try just the PFD client without the CDU.MCP -- fly with the FS autopilot if necessary. See if the gagues perform responsively. Then try the MCP only, using FS defauult instruments to see what you are doing. and so on. You really do need to narrow it down.

To the logs:

Please, it isn't any good sending me PM logs -- they are for PM support. I assume you are talking to them?

- Wideclient logs: as I can see, there are no problems reported in the log (pls see attachment)

What is your average frame rate in FSX by the way. It looks too good to be true for a mere 3GHz PC from the Logs -- averaging 16 frames per second update rates on both clients. Are you only flying in country areas with no autogen? I can't get such good frame rates with the release version of FSX on a Pentium Core2 Extreme 6800!

Have you tried limiting the frame rate in FSX? If you are really achieving 16 average, try limiting it to 15 or 16.

Is FSX your first FS with WideFS and PM? If not, can you make sure it all runs well on FS2004 or FS2002 first?

- Wideserver-log: there are two errors: 10053 and 10054; no idea why this happens.

Now this WideServer log is nothing like the previous one you posted, which showed no errors at all for 2659 seconds (44 minutes) when you closed down. Now you show disconnections and reconnections several times over the course of 3335 seconds (56 minutes), starting after only 8 minutes! Why is this one different to the last? What was different in what you did?

It LOOKS like you simply closed and restarted the Clients several times, hence the disconnections. I say this because the two error-free Client logs you supply start at 17:05, a full 41 minutes after WideServer started, so only relate to the last 15 minutes of the session during which there were no errors logged.

Do you still thnk that this is a PM issue (PM told me that it could be a WideFS-problem

There are no problems shown in the logs, though of course you aren't showing me the whole story by supplying a WideServer log for 56 minutes with Clients relating only to the last 15.

I can only think of three possibilities:

1) Data is getting queued in the TCP/IP link from FSX to SimConnect, and that queue is building up. I don't think this is that likely, but there have been one or two cases reported, early on, with such symptoms -- even on a single machine, not PM. I think these were fixed by uninstalling firewall and security programs and moving to ones known to work okay with FSX -- Windows XP's firewall is okay, and I think AVG and Norton Antivirus are okay. Note that merely disabling them doesn't fix anything, they have to be completely uninstalled in order to remove the hooks that are getting in the way.

2) Data is, for some reason, piling up in the memory buffers of the Fs Pc (the Server) because it isn't being swallowed fast enough across the Network. This could, I suppose, be a wiring or hub problem, or maybe even a Network Card or Network driver problem. Check all of these -- reinstall the Network cards and their drivers (literally if necessary). Try alternative network cards if available. Don't use Wireless.

3) Data is piling up in the Client TCP/IP queues because WideClient isn't getting enough time to process it. Since the PM programs aren't that heavy on processor usage, this is more likely to apply to the PFD PC and be down to graphics drivers taking the processor for large chunks of the time. This has been quite a common cause in the past -- you need video cards and drivers properly supporting OpenGL hardware acceleration and working well. The CDU and MCP don't use OpenGL (well, it is an option on the CDU but I've never used it).

If the problem really is on the PFD PC, and the CDU/MCP work okay without it, then this really is the most likely.

You should discuss these sorts of problems on the PM forums, because there will be folks who had this and dealt with it before. PM used to have their own Newgroups/Forums, but sadly they don't now. Try the group at MyCockpit.

http://www.mycockpit.org/forums/forumdi27981&f=99

Regards

Pete