Multi-class heterogeneous clients in a clustered filesystem
View Patent ↗A cluster of computer system nodes connected by a storage area network include two classes of nodes. The first class of nodes can act as clients or servers, while the other nodes can only be clients. The client-only nodes require much less functionality and can be more easily supported by different operating systems. To minimize the amount of data transmitted during normal operation, the server responsible for maintaining a cluster configuration database repeatedly multicasts the IP address, its incarnation number and the most recent database generation number. Each node stores this information and when a change is detected, each node can request an update of the data needed by that node. A client-only node uses the IP address of the server to connect to the server, to download the information from the cluster database required by the client-only node and to upload local disk connectivity information.
1. A method of operating a cluster filesystem for a cluster connected by a storage area network, comprising:
maintaining cluster configuration data at each client node in a cluster;
repeatedly multicasting from the cluster database a cluster database server address and a database generation identifier;
requesting cluster information from the cluster database server by an existing client node in the cluster in response to the multicast cluster database server address or the database generation identifier being different than the cluster configuration data maintained at the existing client node;
establishing a communication link by a candidate node not a member of the cluster with the cluster database server in response to the multicast cluster database server address and the database generation identifier;
obtaining information from the cluster database server to join the cluster by the candidate node, the information including an address for at least one switch in the storage area network,
obtaining, by the candidate node from the at least one switch, at least one port address of at least one port connected to at least one host bus adapter of the candidate node;
sending addresses of the at least one port and at least one host bus adapter from the candidate node to the cluster database server to update the cluster information concerning the cluster maintained by the cluster database server; and
requesting membership in the cluster by the candidate node.
2. A method as recited in claim 1 , further comprising the candidate node obtaining cluster configuration data in response to requesting membership in the cluster; and
configuring kernel components of the candidate node to enable access to the storage area network by the candidate node upon acceptance as a member of the cluster.
3. A method as recited in claim 1 , further comprising wherein requesting the cluster information includes
comparing the multicast cluster database server address and database generation identifier multicasted by the cluster database server with the cluster configuration data maintained at the existing node of the cluster;
requesting by the existing client node updated cluster configuration data from the cluster database server in response to the multicast cluster database server address or database generation identifier not matching the cluster configuration data maintained at the existing node;
updating the cluster configuration data maintained at the existing node using data returned in response to receipt of the cluster information from the cluster database server; and
configuring kernel processes of the existing client node based on the updated cluster configuration data.
4. A method as recited in claim 3 , further comprising responding to failure and recovery of the cluster database server by detecting a change in at least one of the cluster database server address and the database generation identifier.
5. A method as recited in claim 3 , further comprising performing at least one of a mount and an unmount of a filesystem when the updated cluster configuration data indicates changes affecting the filesystem.
6. A method as recited in claim 1 , further comprising authenticating the candidate node by the cluster database server prior to providing the information required to join the cluster.
7. At least one computer readable medium storing at least one program embodying a method of transmitting a message over a network of nodes running different operating systems, comprising:
maintaining cluster configuration data at each client node in a cluster;
repeatedly multicasting from the cluster database a cluster database server address and a database generation identifier;
requesting cluster information from the cluster database server by an existing client node in the cluster in response to the multicast cluster database server address or the database generation identifier being different than the cluster configuration data maintained at the existing client node;
establishing a communication link by a candidate node not a member of the cluster with the cluster database server in response to the multicast cluster database server address and the database generation identifier;
obtaining information from the cluster database server to join the cluster by the candidate node, the information including an address for at least one switch in the storage area network,
obtaining, by the candidate node from the at least one switch, at least one port address of at least one port connected to at least one host bus adapter of the candidate node;
sending addresses of the at least one port and at least one host bus adapter from the candidate node to the cluster database server address to update the cluster information concerning the cluster maintained by the cluster database server; and
requesting membership in the cluster by the candidate node.
8. At least one computer readable medium as recited in claim 7 , wherein said method further comprises the candidate node
obtaining cluster configuration data in response to requesting membership in the cluster; and
configuring kernel components of the candidate node to enable access to the storage area network by the candidate node upon acceptance as a member of the cluster.
9. At least one computer readable medium as recited in claim 8 , wherein requesting the cluster information includes
comparing the multicast cluster database server address and database generation identifier multicasted by the cluster database server with the cluster configuration data maintained at the existing node of the cluster;
requesting by the existing client node updated cluster configuration data from the cluster database server in response to the multicast cluster database server address or database generation identifier not matching the cluster configuration data maintained at the existing node;
updating the cluster configuration data maintained at the existing node using data returned in response to receipt of the cluster information from the cluster database server; and
configuring kernel processes of the existing client node based on the updated cluster configuration data.
10. At least one computer readable medium as recited in claim 9 , wherein said method further comprises responding to failure and recovery of the cluster database server by detecting a change in at least one of the cluster database server address and the database generation identifier.
11. At least one computer readable medium as recited in claim 9 , further comprising performing at least one of a mount and an unmount of a filesystem when the updated cluster configuration data indicates changes affecting the filesystem.
12. At least one computer readable medium as recited in claim 7 , further comprising authenticating the candidate node by the cluster database server prior to providing the information required to join the cluster.
13. A cluster of computer systems, comprising:
a cluster database server to repeatedly multicast a cluster database server address and a database generation identifier; and
at least one of an existing node in a cluster and a candidate node not in the cluster programmed to request cluster information from the cluster database server to respectively update cluster configuration data maintained by the existing client node in the cluster and obtain information required to join the cluster;
wherein the existing node is operable to request cluster information from the cluster database server in response to the multicast cluster database server address or the database generation identifier being different than the cluster configuration data maintained at the existing client node,
wherein the candidate node is operable to establish a communication link with the cluster database server in response to the multicast cluster database server address and the database generation identifier, obtain an address for at least one switch in the storage area network, and then to obtain from the at least one switch at least one port address of at least one port connected to at least one host bus adapter of the candidate node, to send addresses of the at least one port and at least one host bus adapter to the cluster database server to update the cluster information maintained by the cluster database server, and to request membership in the cluster.
14. A cluster of computer systems as recited in claim 13 , wherein said at least one candidate node is further programmed to obtain cluster configuration data in response to requesting membership in the cluster, and to configure kernel components of the candidate node to enable access to the storage area network by the candidate node upon acceptance as a member of the cluster.
15. A cluster of computer systems as recited in claim 14 , wherein said at least one existing node is programmed to compare the multicast cluster database server address and database generation identifier multicasted by the cluster database server with the cluster configuration data maintained at the existing node of the cluster, to request cluster information from the cluster database server in response to the multicast cluster database server address or the database generation identifier not matching the cluster configuration data maintained at the existing node, to update the cluster configuration data maintained at the existing node using data returned in response to receipt of the cluster information from the cluster database server, and to configure kernel processes of the existing client node based on the updated cluster configuration data.
16. A cluster of computer systems as recited in claim 15 , wherein said at least one existing node is further programmed to respond to failure and recovery of the cluster database server by detecting a change in at least one of the cluster database server address and the database generation identifier.
17. A cluster of computer systems as recited in claim 15 , wherein said at least one existing node is further programmed to perform at least one of a mount and an unmount of a filesystem when the updated cluster configuration data indicates changes affecting the filesystem.
18. A cluster of computer systems as recited in claim 13 , wherein said cluster database server is programmed to authenticate the candidate node by the cluster database server address prior to providing the information required to join the cluster.
19. A cluster of computer systems as recited in claim 13 , wherein the cluster information includes cluster configuration data defining membership of nodes in the cluster.
20. At least one computer readable medium as recited in claim 7 , wherein the cluster information includes cluster configuration data defining membership of the nodes in the cluster.
21. A method as recited in claim 1 , wherein the cluster information includes cluster configuration data defining membership of nodes in the cluster.