All Nodes Down Alert Message (SNMP?)

I don't see an alert or existing log message you could leverage, either.

You could create a rule that logs a custom message, then create a custom alert for it:

Example rule:

when CLIENT_ACCEPTED {
   if { [active_members myPool] < 1 } {
      log local0. "No nodes available for virtual server [virtual name] at [IP::local_addr]:[TCP::local_port]"
      reject
   }
}

I'll leave it to you to figure out the custom alert definition & mapping if you think this might work for you.

HTH

/deb

Wow, that was easy. Your suggestion worked like a champ. I'm getting tons of repeat log messages, however (only one email, fotunately). Can I limit the number of log messages? Thanks, Ken

You could use the clock command to log the condition on the minute, or whatever interval you choose.

when CLIENT_ACCEPTED {
  if { [active_members myPool] < 1 } {
    if { [clock format [clock seconds] -format {%S} ] eq "00" } {
      log local0. "No nodes available for virtual server [virtual name] at [IP::local_addr]:[TCP::local_port]"
    }
    reject
  }
}

updated to move the reject statement out of the log condition.

You might also just try removing the local0. reference & use this log line instead:

 log "No nodes available for virtual server [virtual name] at [IP::local_addr]:[TCP::local_port]"

Without the specific destination reference, you'll probably see the syslog summary messages instead:No nodes available for virtual server myVirtual at 1.1.1.1:80

Last message repeated 42,1764 times.

/d

Removing local0 didn't make any difference. Thanks, Ken

How could I check for no nodes available for, say, a period of 5 minutes, rather than immediately? Thanks, Ken