Replicating the Internet Lockout database

As an administrator, you need to decide whether replicating the Internet Lockout database to other servers is useful for you.

About this task

A major advantage of replicating the database is that lockout information is replicated to multiple servers. You can look at any replica and determine the lockout status of users across multiple servers, instead of having to open up the Internet Lockout database on each server for which Internet password locking is enabled.

However, replication also has its disadvantages; for example, replication storms can occur if your network is under attack, or if a denial-of-service attack occurs. Additionally, if replication is slow to occur, anyone checking the lockout database on a given server might not be able to see that a person is locked until replication occurs (however, they can always open the replica directly on the server in question).

The Internet Lockout database is created with a replica ID that stays the same for any replica on any server for which Internet password locking is enabled in a domain. By default, replication is temporarily disabled for Internet lockout databases. This is to prevent replication storms described earlier. To replicate the database to another server, disable the Temporarily disable replication option in the Other section of the Replication Settings dialog box. You can then set up the database to replicate (either scheduled or clustered replication).

Note: When you replicate this database to other servers, the 'invalid attempts' information is calculated for each individual server. For example, if the threshold for 'John Doe' is three, and he has two invalid attempts on Server A and has one on Server B, he is not locked out of either server. The attempts are not combined for a total of three. The reason for replication is ease of administration, not to establish global thresholds.

Procedure