Method, system and program products for managing processing groups of a distributed computing environment
First Claim
Patent Images
1. A method of managing processing groups of a shared nothing distributed computing environment, said method comprising:
- requesting via a request by a prospective member to join a processing group of a shared nothing distributed computing environment, said request including a sequence number indicating a version of the processing group, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk;
determining whether the prospective member can join the processing group, said determining employing the sequence number, wherein the determining comprises comparing by said prospective member the sequence number in the request with a current group sequence number to determine if the join of the prospective member to the processing group should continue;
joining the processing group by the prospective member, in response at least in part to the determining indicating that the prospective member can join the processing group; and
wherein the joining comprises automatically reinitializing state of the prospective member responsive to the comparing indicating that the sequence number in the request is less than the current group sequence number, the reinitializing making a state of the prospective member consistent with a state of an existing member of the processing group, and thereafter, proceeding with the joining by the prospective member.
3 Assignments
0 Petitions
Accused Products
Abstract
The management and use of replicated distributed transactions is facilitated. A distributed synchronous transaction system protocol is provided to manage the replication of distributed transactions for client application instances. The distributed synchronous transaction system allows transactions to be replicated without having the client application instances be aware of other instances to receive the transaction. Further, if a failure occurs during processing of a distributed replicated transaction, the distributed synchronous transaction system manages the recovery of the failure.
39 Citations
29 Claims
-
1. A method of managing processing groups of a shared nothing distributed computing environment, said method comprising:
-
requesting via a request by a prospective member to join a processing group of a shared nothing distributed computing environment, said request including a sequence number indicating a version of the processing group, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; determining whether the prospective member can join the processing group, said determining employing the sequence number, wherein the determining comprises comparing by said prospective member the sequence number in the request with a current group sequence number to determine if the join of the prospective member to the processing group should continue; joining the processing group by the prospective member, in response at least in part to the determining indicating that the prospective member can join the processing group; and wherein the joining comprises automatically reinitializing state of the prospective member responsive to the comparing indicating that the sequence number in the request is less than the current group sequence number, the reinitializing making a state of the prospective member consistent with a state of an existing member of the processing group, and thereafter, proceeding with the joining by the prospective member. - View Dependent Claims (2, 3, 4, 5, 6)
-
-
7. A method of managing processing groups of a shared nothing distributed computing environment, said method comprising:
-
joining a prospective member to an inactive processing group of a shared nothing distributed computing environment, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; comparing a sequence number of the processing group with a sequence number of the prospective member; updating the sequence number of the processing group, in response to the comparing indicating a particular difference; determining whether a quorum of members has joined the processing group; setting the sequence number of the processing group, in response to the determining indicating a quorum of members has joined the processing group; initiating activation of the processing group, in response to the setting; and wherein the initiating activation comprises; obtaining by a member of the processing group having a sequence number lower than the sequence number of the processing group a copy of group state associated with the sequence number of the processing group; and reinitializing the member using the copy of group state. - View Dependent Claims (8, 9, 10)
-
-
11. A system of managing processing groups of a shared nothing distributed computing environment, said system comprising:
-
a request by a prospective member to join a processing group of a shared nothing distributed computing environment, said request including a sequence number indicating a version of the processing group, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; means for determining whether the prospective member can join the processing group, said means for determining employing the sequence number, wherein the means for determining comprises means for comparing by said prospective member the sequence number in the request with a current group sequence number to determine if the join of the prospective member to the processing group should continue; means for joining the processing group by the prospective member, in response at least in part to the determining indicating that the prospective member can join the processing group; and wherein the means for joining comprises means for automatically reinitializing state of the prospective member responsive to the means for comparing indicating that the sequence number in the request is less than the current group sequence number, the reinitializing making a state of the prospective member consistent with a state of an existing member of the processing group, and thereafter, proceeding with the joining by the prospective member. - View Dependent Claims (12, 13, 14, 15, 16)
-
-
17. A system of managing processing groups of a shared nothing distributed computing environment, said system comprising:
-
means for joining a prospective member to an inactive processing group of a shared nothing distributed computing environment, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; means for comparing a sequence number of the processing group with a sequence number of the prospective member; means for updating the sequence number of the processing group, in response to the comparing indicating a particular difference; means for determining whether a quorum of members has joined the processing group; means for setting the sequence number of the processing group, in response to the determining indicating a quorum of members has joined the processing group; means for initiating activation of the processing group, in response to the setting; and wherein the means for initiating activation comprises; means for obtaining by a member of the processing group having a sequence number lower than the sequence number of the processing group a copy of group state associated with the sequence number of the processing group; and means for reinitializing the member using the copy of group state. - View Dependent Claims (18, 19)
-
-
20. An article of manufacture comprising:
-
at least one computer usable storage medium having computer readable program code logic stored thereon to manage processing groups of a shared nothing distributed computer environment, the computer readable program code logic comprising; a request by a prospective member to join a processing group of a shared nothing distributed computing environment, said request including a sequence number indicating a version of the processing group, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; determine logic to determine whether the prospective member can join the processing group, said determining employing the sequence number, wherein the determining comprises comparing by said prospective member the sequence number in the request with a current group sequence number to determine if the join of the prospective member to the processing group should continue; join logic to join the processing group by the prospective member, in response at least in part to the determining indicating that the prospective member can join the processing group; and wherein the join logic comprises automatically reinitialize logic to automatically reinitialize state of the prospective member responsive to the comparing indicating that the sequence number in the request is less than the current group sequence number, the automatic reinitialize logic making a state of the prospective member consistent with a state of an existing member of the processing group, and thereafter, proceeding with the joining by the prospective member. - View Dependent Claims (21, 22, 23, 24, 25)
-
-
26. An article of manufacture comprising:
-
at least one computer usable storage medium having computer readable program code logic stored thereon to manage processing groups of a shared nothing distributed computer environment, the computer readable program code logic comprising; join logic to join a prospective member to an inactive processing group of a shared nothing distributed computing environment, wherein the shared nothing distributed computing environment is an environment with a particular system architecture wherein only one processing node is connected to a given disk; compare logic to compare a sequence number of the processing group with a sequence number of the prospective member; update logic to update the sequence number of the processing group, in response to the comparing indicating a particular difference; determine logic to determine whether a quorum of members has joined the processing group; set logic to set the sequence number of the processing group, in response to the determining indicating a quorum of members has joined the processing group; initiate logic to initiate activation of the processing group, in response to the setting; and wherein the initiate logic comprises; obtain logic to obtain by a member of the processing group having a sequence number lower than the sequence number of the processing group a copy of group state associated with the sequence number of the processing group; and reinitialize logic to reinitialize the member using the copy of group state. - View Dependent Claims (27, 28, 29)
-
Specification