You need to change the fail-over switch type. Currently you have it configured so when the Standby takes over it doesn't activate the remote-control features and so it doesn't let other schedulers to connect to it. That's what the auto-pilot mode is all about. What you want is a cyclying mode in wich Standby becomes a new Master after the fail-over event. Here is a link to the page describing how to do that
http://www.softtreetech.com/cgi_bin/config.cgi?read=7379 : We're testing a fail-over situation with two servers
: pointing to each other like this: Server1: Distributed Server Mode: Enabled
: Sync Interval: 30
: Fails before takeover: 1
: Driver: WinSock
: Port: 10000
: Location: Server2 : Server2: Distributed Server Mode: Enabled
: Sync Interval: 30
: Fails before takeover: 1
: Driver: WinSock
: Port: 10000
: Location: Server1 : We started Server1, and it worked OK. We started Server2; it went into
: fail-over mode. Fine.
: We tried a Remote Connection to Server1: that worked. A Remote Connection to
: Server2 did not work (not active). That seems normal.
: Then we stopped the Scheduler on Server1. Server2 polled 3x and took over
: (auto-pilot mode).
: At this time we needed a remote 24x7 connection to Server2 but still couldn't
: get one. And when we started Server1 again, it entered Master mode and all
: jobs were executed twice!
: Only a re-start of the scheduler on Server2 would help; but we need a remote
: connection with the fail-over server without a re-start.
: We tried the same thing with NamedPipes, but with the same result.
: Can you help us? : Thanks! : Ekke
: (not Jason)
|
