Restrictions – Kofax Capture User Manual
Page 22
Installation
22
Kofax Reporting Administrator's Guide
b Specify an IP address of one of the cluster hosts and click Connect. Select an interface
and click Next.
c It is recommended to change the Default state to Suspended and to select "Retain
suspended state after computer restarts". Message Connector updates the state after
startup. Click Next.
d Click Add and specify the cluster IP address. Click OK, then click Next.
e In the Full Internet name field, enter the name of the cluster computer.
f Select Multicast (unless you have at least 2 network adapters on each node). Click
Next.
g Click Add to add port rules for web service input (default 25500 for receiver instance
1). Select Multiple host as a Filtering mode and set Affinity to None.
h Click Finish.
5 Add additional cluster nodes:
a Right-click the cluster and select Add Host.
b Specify an IP address of the node and click Connect. Select an interface and click
Next.
c Click Next and Finish.
Restrictions
The NLB port control still leaves some short unavailability time window (usually 1 to 8
seconds) when the failover is not provided and a web service connection to the common NLB
IP address fails. This occurs (always) in the following situations:
▪ Unavailability time window when WSA node stops abruptly: In this case it is the NLB (and
not the port control) that handles this situation. NLB requires ~5 seconds to detect a failed
node. NLB requires 2 to 3 seconds to remove the failed node and redistribute its load to the
live nodes.
▪ Unavailability time window when a node starts. Some seconds elapse between when
the NLB service starts (with enabled ports - depends on cluster host configuration) and
when the WSA with port control starts. You can eliminate most of this delay by properly
configuring the cluster: refer to
, step 4c.
▪ Unavailability time window at WSA process crash. Some (1…3) seconds elapse until the
WSA port control detects the crash and the disables the ports for NLB.
Nevertheless, if the web service client supports retries, the second attempt will presumably be
outside of this time window and the connection request will be directed to a working node.