I run a FSMA 12/12 FMPA on a newly configured logical Windows 2008 R2 server.
The FMPA12 client runs every two seconds a script on the server that looks whether a PDF report must be created.
This FMPA12 client, although constantly active, is disconnected every one hour or so.

 

The FSMA 12 log shows: client .... no longer responding, connection closed (51).
This issue is regularly discussed in several forums, but FileMaker as usual never comes up with a solution.

Exactly the same functionality running on another logical Windows2008 R2 server for years without problems.
Does anyone know a solution to this problem?

Thanks in advance

Not to be too obvious, but does the client machine sleep after an interval?

Thanks but client machine sleep settings are on 'never'.
Also clients settings in privilegesets are such that clients are not to be disconnected ever, even not if idle.

Any robot service is also liable to actual physical network issues that cause disconnection.

 

I tend to run a scheduled task that opens the relevant file, runs the process and shuts down again.

 

Every 2 sec seems excessive, but you could vary this to run once an hour, with the 2 sec script set to run 1780 times then close down... to give time for the close down to occur before the next instance of the scheduled task.

The client is supposed to send the host a "tickle" if there has been no other network activity to the host for 60 seconds.

 

If the host doesn't hear from a client for 125 seconds and doesn't have anything to send to the client, it will disconnect the client since this means the client has failed to send two "tickle" messages. 

 

51 means "Client tickled out", the 125 second timeout was reached with no contact from client.

 

So if you're seeing 51 errors, it means the client is trying to send a "hello, I'm still here" tickle, but it's not being received by the server, so the server eventually disconnects the client after 125 seconds. 

 

This is probably being caused by a network configuration issue.  So check NIC's, routers, switches, etc.

 

Steven

Hi, I hope we found the cause and the solution as well.

The server has 2 NIC's, one for RDP/VPN connections to the server and one for standard FM data traffic.
The NIC for RDP/VPN was set als primary.

 

This probably causes the network services to lose connection with the FMPA12 client.
We have now set the second NIC as primary and hope this will solve the problem.
We have to wait for about 2 hours and see. I will keep you updated.

Was this ever resolved?

 

I have the same issue. Though my clients that are getting kicked off are at a remote location and are accessing the Server databases via an internet connection that leaves something to be desired (6Mbps down/1.5Mbps up).

 

Mr. Blackwell, is there any way around the 125 second timeout setting?

Sorry for delayed response.  I do not know know of any way around the 125 second item.  It's there intentionally.

 

Steven

Thanks, Steven.

 

I am actually getting satisfactory results now that I had my remote clients take their computer sleep and hard drive sleep settings down to Never. They are reporting that they almost never get kicked off anymore. And they shut down every night, so that's not an issue.

That may then have been the proximate cause of their failure to respond to the tickles.  I am pleased to hear things are working better now.

 

Steven

Does anyone know the difference between a:

 

no longer responding; connection closed. (10)

 

no longer responding; connection closed. (51)

 

?

The client is supposed to send the host a "tickle" if there has been no other network activity to the host for 60 seconds.

 

If the host doesn't hear from a client for 125 seconds and doesn't have anything to send to the client, it will disconnect the client since this means the client has failed to send two "tickle" messages.

 

As for the errors: FileMaker Server will log a message like one of the following in the Access.log file when a client connection fails, assuming Client Access logging is enabled:

 

[Client Name, etc] no longer responding; connection closed. (10)

[Client Name, etc] no longer responding; connection closed. (51)

 

The (10) indicates that the 125 second timeout elapsed as described above. The (51) indicates that the TCP connection failed when Server tried to send something to the client. 

 

Steven

Filemaker hung - a lot like it did before we installed the 11.0.5.510 server upgrade, which we did long ago but are still on.  We had multiple connections for the same person in the admin console and no one could be closed.  Users just got an hourglass  Backup failed due to lock.  The only solution seemed to be rebooting the server.

 

Prior to the incident the event.log shows that everyone (so it seems) was no longer responding; connection closed. (10) with a few (51).  While 51 is described above as being network card/config related....The 10's may have been due to the ghost logins as the same user shows multiple times within the same minute. But while the log says they were closed, they still appeared in the console.

 

Any idea what would have caused this incident?