Technical Tip: Virtual servers load balancing showing wrong status when failover within a short time
Description
This article describes how virtual servers load balancing showing wrong status when failover within a short time.
Solution
This is an expected behavior:
In the active-standby mode, if the primary server went down, FortiOS will not forward any traffic to it for 5 minutes based on the holddown_interval default value.
This value prevents any flapping issues (if a server goes up and down frequently).
This article describes how virtual servers load balancing showing wrong status when failover within a short time.
Solution
This is an expected behavior:
In the active-standby mode, if the primary server went down, FortiOS will not forward any traffic to it for 5 minutes based on the holddown_interval default value.
This value prevents any flapping issues (if a server goes up and down frequently).
Server A as master: x.x.x.xFrom the 1 up status and 1 down status within the load balance monitor.
Server B as standby: y.y.y.y
x.x.x.x:80 Up Active 4 0 <1 second(s) 0 BWhen Server A is down:
y.y.y.y:80 Down Standby 1 0 <1 second(s) 0 B
x.x.x.x:80 Down Active 5 0 <1 second(s) 0 BWhen Server A bring up (within the holddown_interval :(
y.y.y.y:80 Up Active 2 0 <1 second(s) 0 B
x.x.x.x:80 Down Active 8 0 <1 second(s) 0 BIf Server A is brought up after holddown_interval :
y.y.y.y:80 Up Active 4 0 <1 second(s) 0 B
x.x.x.x:80 Up Active 6 0 <1 second(s) 0 BIt is possible to lower the holddown_interval for the testing (default is 5 minutes, lower it to 30 second the least).
y.y.y.y:80 Down Standby 3 0 <1 second(s) 0 B
# config firewall vip
edit <vserver>
# config realservers
edit 1
set holddown-interval <x> <----- Where x is the seconds for the hold down.
next
edit 2
set holddown-interval <x> <----- Where x is the seconds for the hold down.
next
end
