Journalling optimization system and method for distributed computations

• US 5,371,889 A
• Filed: 04/22/1993
• Issued: 12/06/1994
• Est. Priority Date: 04/22/1993
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: Klein, Johannes
• Primary Examiner(s): Kriess, Kevin A.
• Assistant Examiner(s): KATBAB, ABDOLLAH
• Application Number:
  US08/051,523
• Time in Patent Office: 593 Days
• Field of Search: 395/650, 364/DIG. 1, 364/285, 364/285.1, 364/285.2, 364/285.3
• US Class Current: 718/106
• CPC Class Codes: G06F 9/466 Transaction processing

Abstract

A protocol analysis system is provided with data specifying the defined states of processes participating in a distributed computation. State transitions between states are specified as being enabled by (A) receiving a message, (B) unreliably sending a message, or (C) performing an external action such as reliably sending a message. The specification data also identifies process states known to be final states, and all other states are initially denoted as intermediate states. The protocol analysis system determines if any intermediate states can be re-categorized as final states. Then it determines if any state transitions initially identified as unreliable send operations must be treated as derived external actions, and thus made reliable. Thirdly, for each derived external action, the states of the affected application process must be re-evaluated so as to determine if derived final states need to be converted into intermediate states. The resulting determinations as to which states are final states and which messages must be reliable sent are recorded and used to govern execution of the application process. When executing the application process, state transitions entering and leaving intermediate states are normally recorded on stable storage before the state transition is carried out and reliably sent messages are normally recorded on stable storage before being sent. A number of run-time journal optimization techniques reduce the number of state transitions and messages that need to be stored on stable storage.

Claims

1. In a computer system executing application processes that interactively perform a distributed computation, a method performed by said computer system of executing the distributed computation, the method comprising:

identifying, using a first set of journalling criteria, for each application process first states of said application process that require journalling of state information and events causing a transition from one of said first states to another state of said application process, said first set of journalling criteria providing criteria for synchronous journalling wherein a message is sent to indicate a transition into an intermediate state after receiving confirmation that said journalling has been successfully completed.

identifying, while executing each application process, and using a second set of journalling criteria that supersede said first set of journalling criteria, second states of each application process that can be journalled asynchronously and events enabling transitions between ones of said states that can be journalled asynchronously, said second set of journalling criteria utilizing state and event information determined during said executing; and

synchronously journalling, while executing each application process, said first states of said application process which have not been determined as being one of said second states; and

asynchronously journalling said second states of said application process.

2. The method of claim 1 further comprising:

identifying while executing each application process a sequence of states using predefined criteria, said sequence comprising a first state and a last state and said predefined criteria specifying when journalling may be avoided for all other states of said sequence; and

journalling while executing said application process only said first and last states of said identified sequence of states.

3. The method of claim 2 wherein said sequence of states includes at least one state transition enabled by reliably sending a message, and the method further includes delaying journalling of said last state until receipt of said reliably sent message has been acknowledged by its recipient.

4. The method of claim 3

wherein said sequence of states includes at least one state transition caused by receiving a reliably sent message from another application process, and the method further includes delaying sending an acknowledgement of receiving said reliably sent message to said other application until said last state has been journalled.

6. The method of claim 1, wherein each state of said sequence of states of each application process is one of an initial state representing a start state of said application process, an intermediate state representing a state from which said application process is unable to terminate its execution, or a final state representing a state from which said application process is able to terminate its execution.

7. The method of claim 6, wherein said first set of journalling criteria comprises journalling to stable storage intermediate state information, final state information for a first final state entered after a sequence of intermediate states, and a message used for process communication prior to being sent if said message is sent repeatedly until a receiver process has acknowledged receipt of said message.

8. The method of claim 1, wherein each of said events causing a transition comprises one of receiving a message, reliably sending a message until received by another process, or unreliably sending a message which is not guaranteed to be received by another process.

9. The method of claim 1, wherein a first of said application processes shares a common journal with another of said application processes, and said first of said application processes uses asynchronous journalling.

10. The method of claim 2, wherein said sequence of states contains no conditional branches from a state belonging to said sequence.

11. The method of claim 2, wherein said predefined criteria uses runtime information obtained during said executing to determine when journalling may be avoided.

5. In a computer system executing processes that interactively perform a distributed computation, a method performed by said computer system of executing the distributed computation, the method comprising:

identifying using a first set of journalling criteria, for each application process first states of said application process that require journalling of state information and events causing a transition from one of said first states to another state of said application process, said first set of journalling criteria providing criteria for synchronous journalling wherein a message is sent to indicate a transition into an intermediate state after receiving confirmation that said journalling has been successfully completed

identifying, while executing each application, using a second set of journalling criteria that supersede said first set of journalling criteria, a sequence of states starting with a first final state and having only one state transition enabled by a reliable send operation and ending in a second final state, said second set of journalling criteria utilizing state and event information to identify said sequence of states that may be reordered to avoid journalling;

reordering, while executing each application process, said sequence of states to produce a reordered sequence of states such that said reliable send operation is executed first and other state transition events are executed later, and when executing said reordered sequence of states avoiding journalling of states and state transitions which follow said first identified final state in said reordered sequence, and delaying execution of said other state transition events until said one reliable send operation is acknowledged by a recipient.

12. The method of claim 5, wherein said second set of journalling criteria uses runtime information obtained during said executing to identify said sequence of states that may be reordered to avoid journalling.