web companion crashes daily

[mlindal]

So for those of you following our plight....

We upgraded our FM server (W2K) to ver 5.5 and a new machine at 1.8GHz.

Still connecting Web companion on an iMac (500Mhz) under Jaguar 10.2.1 - all reinstalls and updates done. Clean install of all OS's, Jupiter and Saturn are far enough apart...

Every morning we find the web companion froze or FM 5.5 unlimited on the Mac crashed (crash log says <EXC_BAD_ACCESS (0X0001)> AND <KERN_INVALID_ADDRESS (0X0001) AT 0X06376000>

Thanks to all who have helped so far - but one year later - still flakey.

What is next? WSC? Lasso? JSP?

[Garry Claridge]

Are you using 5.5v2 of WebCompanion? What about Venus and Mars?

Garry

[mlindal]

Web companion is up to date - all patches done. I forgot about Venus and Mars... Maybe it is because the water flows backwards down the drain as opposed to Australia.

[Anatoli]

Try to do the opposite, server on Mac and Unlimited on PC. Will that work better?

We are running NT4 server with Unlimited from May 25th without restart.

[mlindal]

We tried opening the files locally on the web companion machine. (ie no server involved) to see if it was a network problem. It crashed again. We are now going back to the server today.

We are now looking at cdml code to see if there is something in that that is causing FM to crash. Can't think of what - but some consultant will make some money helping us...

[Anatoli]

BTW -- rock solid is Unlimited 5. IMHO Not the 5.5.

[mlindal]

Yes - I think you are right... but we are stuck with 5.5 for now under OS 10.2. However - what about FM 6 unlimited? would that be better?

[Garry Claridge]

Have you increased the Filmaker Cache size to, at least, 8192KB. I'm just guessing here. Do you have Sun-Spot protection enabled

Good Luck.

Garry

[mlindal]

Cache is maxed out - under OS 10.2 shouldn't matter much.

Re: sun spots - you are not far off Garry, we now have a full firewall on the Imac and blocked a few users that were hitting us 800 times a day. We also found flakey connected to the smtp server (remote) and that one of the local windows servers was trying to do something on port 139. We also have a cpu monitor running - and at times of crash it goes to 100% and stays there. Since we blocked port 139, it is well below half.

One day without crashing and counting...

[mlindal]

looks like things are solved. The firewall keeps the bad guys out and the cpu monitor is constant and Filemaker is happy. The planets are back in alignment and the good Lord is shining upon us again.

[Garry Claridge]

Yahoo!

Garry

[Anatoli]

BTW 139 port relates to:

The NetBIOS name acts as a session layer application identifier. For example, the NetBIOS Session service operates over TCP port 139. All NetBIOS over TCP/IP session requests will be addressed to TCP destination port 139. To identify the NetBIOS application with which to establish a NetBIOS session, the NetBIOS name is used.