Cluster-free techniques for enabling a directory protocol-based domain name system (DNS) service for high availability
View Patent ↗Cluster-free techniques for enabling a directory protocol-based Domain Name System (DNS) service for high availability are presented. A DNS service monitors a node for wild-carded IP address that migrate to the node when a primary node fails to service DNS requests for a directory of the network. The DNS service forwards the wild-carded IP address to a distributed directory service for resolution and uses the distributed directory service to dynamically configure the DNS service for directly handling subsequent DNS requests made to the directory over the network while the primary node remains inoperable over the network.
1. A method implemented in a non-transitory computer-readable storage medium to process on a node and to perform the method, comprising:
listening, on the node, for a wild-carded Internet Protocol (IP) address and recognizing the wild-carded IP address as a Highly Available IP (HaIP) address that is bound to a network resource and decoupled from a physical interface, wherein the HaIP address migrates to the node when an original node that was used to resolve the HaIP address for original DNS requests fails to respond;
receiving, at the node, a first Domain Name System (DNS) request; and
communicating from the node and over a network with a directory service for a directory maintained by the directory service, communication with the directory service indicates that the first DNS request was received on the wild-carded IP address, the first DNS request is asked to be serviced from the directory service that maintains DNS data for the directory, and wherein the directory service is external to the network node over the network, and wherein subsequent DNS requests are serviced from the node;
acquiring, on the node, the DNS data from the directory service; and
rebuilding, on the node, a directory tree for the directory using the DNS data to directly service from the node the subsequent DNS requests node.
2. The method of claim 1 , wherein acquiring further includes, downloading a cache state and cache hit ratio for the directory with the DNS data from the directory service over the network.
3. The method of claim 2 , wherein downloading further includes storing the DNS data, the cache state, and the cache hit ration on a local disk of the node.
4. The method of claim 3 , wherein rebuilding further includes loading from the local disk the cache sate and the cache hit ratio to memory of the node when rebuilding the directory tree.
5. The method of claim 1 , further comprising, configuring the node to begin listing for a specific IP address associated with the wild-carded IP address for directly servicing the subsequent DNS requests from the node.
6. The method of claim 1 , wherein listening further includes periodically receiving configuration updates over the network for the directory from the directory service at predefined intervals, and wherein at least some updates occur before the wild-carded IP address is detected as being received on the node.
7. A system, the system comprising:
a first device having a distributed directory service stored in a non-transitory computer-readable storage medium and to execute on the first device of a network;
a recovery node having a Domain Name System (DNS) service stored in a non-transitory computer-readable storage medium and to execute on the recovery node of the network;
wherein the distributed directory service is to monitor a directory being managed on a primary node of the network, and wherein the DNS service is to forward a wild-carded Internet Protocol (IP) address that migrates dynamically to the recovery node when the primary node fails, and wherein the distributed directory service is to resolve the wild-carded IP address for the DNS service and facilitate configuring the recovery node and the DNS service for resolving subsequent DNS requests made over the network to the directory and wherein the wild-carded IP address is a highly available IP (HaIP) address that is bound to a network resource of the directory and is decoupled from any specific physical device of the network and the HaIP address migrates to the recovery node when the primary node that was used to resolve the HaIP address for original DNS requests fails to respond, and wherein DNS data is acquired on the recovery node from the distributed directory service and a directory tree is rebuilt for the directory using the DNS data to directly service from the recover node the subsequent DNS requests.
8. The system of claim 7 , wherein the distributed directory service is to monitor the directory by maintaining the DNS data for the directory as the directory is accessed and modified on the primary node and while the primary node is operational.
9. The system of claim 8 , wherein the DNS data includes a last cache state and a last cache hit ration that existed on the primary node before it failed.
10. The system of claim 9 , wherein the DNS service downloads the DNS data from the distributed directory service and the first device to a local disk of the recovery node.
11. The system of claim 10 , wherein the DNS service uploads the last cache state and the last cache hit ratio to memory of the recovery node and uses the DNS data to rebuild a directory tree for the directory for purposes of directly resolving the subsequent DNS requests without assistance from the distributed directory service.