Parameters in the Centrify configuration file control many aspects of Centrify DNS subsystem operation. Although you can set any of these parameters, the default settings should provide you with optimal DNS operation. See the Configuration and Tuning Reference Guide for details about any of these parameters.
The DNS subsystem periodically checks in the background to see if a DNS server that is faster than the currently selected one is available. The dns.alive.resweep.interval parameter determines how often this background check occurs; the default value is one hour (3600 seconds).
When a DNS server is selected, its address is stored in the kset.dns.server file, and it is used for all DNS requests until one of the following occurs:
- The selected server stops responding.
- A new server sweep discovers a faster DNS server and replaces it.
- The adclient process is stopped and restarted, which triggers a sweep for a new DNS server.
- The specified server is no longer in the list of servers in /etc/resolv.conf.
For the sweep, the dns.sweep.pattern parameter determines the probe pattern that is used to find a live DNS server; that is, it sets the protocol to use (TCP or UDP) and the amount of time to wait for a response. By default, this parameter specifies both a TCP and UDP probe.
The dns.timeout and dns.udp.retries parameters determine the amount of time to wait, and how often to re-send a request when the current server does not respond to a request. If the current server does not respond to a request within the specified time out period, it is considered down and Centrify looks for a different server. If it cannot find a live server, DNS is considered down, and the Centrify Agent for *NIX waits for the period of the dns.dead.resweep.interval parameter, 60 seconds by default, before performing a sweep to find a new server.