|
I'm looking into the Server Initiated Sync (SIS) for SQL Anywhere 12. I followed the (Tutorial: Server-intiated synchronization using light weight polling). When I follow the steps on my local machine it works. My second step was to try it out on our test server. The consolidated database, notifier and mobilink server I set up on our test server, and the remote database and listener on my computer. That seems to be connecting okay, the listener checks for new messages and starts the sync. But then while sync'ing, the mobilink client always fails with these errors E. Unable to connect a socket. The system-specific error code is 10061 (hex 0000274d).
I found this error (MobiLink Server 10061) but I don't know if it is the system-specific error noted above. Does anybody know what the above errors mean? Like I said earlier the tutorial works when I point it at my local computer, just not the test server. |
|
I suspect that you've defined an action for a message to be to fire off dbmlsync, but the connection information you've put in the remote database to synchronize is still looking for a MobiLink Server on localhost, and not the machine where the MobiLink Server is running. If you look in the SYS.SYSSYNC table for the remote database, does the "server_connect" column for the row with non-null values for "publication_id" and "site_name" contain the string "host=localhost"? If so, you'll need to change the CREATE SYNCHRONIZATION SUBSCRIPTION command you use to direct dbmlsync to the host on which the ML Server is running. I wanted to put ten dollars on "Reg to win" but I couldn't find anyone to bet against me :)
(06 Mar '12, 16:38)
Breck Carter
Replies hidden
Yep, one of SQL Anywhere's many reasonable "defaults":)
(07 Mar '12, 03:50)
Volker Barth
|
10061 may be the most common TCPIP error code... it means "connection refused"... the server you asked for isn't running at the location you sent the message to: http://support.ipswitch.com/kb/WSK-19980709-EM01.htm
MobiLink error codes are negative, so the error you link to (-10061) is not the same as 10061. See Breck's response for the correct interpretation.