Method for managing reuse of hard log space by mapping log data during state changes and discarding the log data

• US 4,878,167 A
• Filed: 06/30/1986
• Issued: 10/31/1989
• Est. Priority Date: 06/30/1986
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Rader, Holly A., Strickland, Jimmy P., Kapulka, Kenneth M.
• Primary Examiner(s): Williams, Jr., Archie E.
• Assistant Examiner(s): Lee, Thomas L.
• Application Number:
  US06/880,387
• Time in Patent Office: 1,219 Days
• Field of Search: 364/200 MS File, 364/900 MS File, 364/300, 371/12, 371/13
• US Class Current: 714/16
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery
• CPC Class Codes: G06F 16/2358 Change logging, detection, and notification replication G06F16/27
• CPC Class Codes: G06F 16/2379 Updates performed during online database operations; commit processing
• CPC Class Codes: G06F 2201/87 Monitoring of transactions

Abstract

A method for writing tagged (partitioned and classified) records from a first log stream to multiple recovery streams and discarding same from said first stream at the termination of the unit of recovery in a transaction-oriented system to permit first log stream reuse.

Claims

1. A method, implemented on a digital computer system, for managing the reuse of space on a hard log stored on memory means in said computer system in support of forward resource recovery (FIG. 4) in a transaction-oriented system in which concurrently executing, failure-independent processes share a common hard log and one or more resources, comprising the steps of:

(a) recording log data on said memory means with a log manager in said computer system in a partitioned fashion by a unit of recovery (UR) in a transaction recovery stream (TRS) resident on the hard log, said UR selectively assuming an inflight, ensure commit, or ensure abort state (FIGS. 2, 3,

     6);

(b) classifying using said log manager to classify the log data into subsets (BEGIN, COMMIT, ABORT, END, REDO, UNDO) according to the function of their role in recovery (FIGS. 6,

     8);

(c) responsive to UR state changes from either inflight to ensure commit (FIG.

     2) or inflight to ensure abort (FIG.

     3), using said log manager to perform a one-to-n mapping of the log data of one of said subsets (typed REDO or UNDO shown in FIGS. 9, 10, 13-16) from a given UR partition on the TRS into resource recovery streams (RRSs) (FIGS. 5,

     7); and

(d) at the termination (END) of the UR state changes utilizing the RRSs, using said log manager to discard all log data identified on the hard log as belonging to the given UR partition from the TRS rendering space occupied by said discarded log data on said hard log amenable to reuse (FIGS. 11, 17-24).

2. The method according to claim 1, wherein the recorded log data in the TRS includes data to determine the UR state, data necessary to ensure an abort of the UR, and data necessary to ensure a commit of the UR.

3. The method according to claim 1, wherein the subset of log data subject to a one-to-n mapping includes records classified as REDO when the UR state is to commit (FIGS. 15, 16) or UNDO when the UR state is to abort (FIGS. 13, 14).

5. The method according to claim 1 or 4, wherein a UR partition defined on each of m resources is distributed among n RRSs where m>

n such that at least one RRS contains the log data from two resources (FIGS. 18, 20, 22,

     24).

4. A method, implemented on a digital computer system, for managing the reuse of space on a hard log stored on memory means in said computer system in support of forward resource recovery (FIG. 4) in a transaction-oriented system in which concurrently executing, failure-independent processes share a common hard log and one or more resources, and further wherein each transaction utilizes begin, commit, abort, redo/undo operations to bound the transaction, and in which redos ensure transaction return to the most recent commit point and undos ensure return to the transaction begin point, comprising the steps of:

(a) initiating the recording of log data on said memory means with a log manager in said computer system in a partitioned fashion by a unit of recovery (UR) using a begin marker on a transaction recovery stream (TRS) resident on the hard log (FIG.

     8);

(b) using said log manager to classify and then to write and then writing to the TRS log data into subsets according to the COMMIT, ABORT, REDO, UNDO function of their role in recovery (FIGS. 6, 8,

     12);

(c) upon a UR state change responsive to a commit operation, used said log manager to perform a one-to-n mapping of all log data in the REDO subset of the UR partition from the TRS into predeterminedly associated resource recovery streams (RRSs), and upon a UR state change responsive to an abort operation, to perform a one-to-n mapping of log data in the UNDO subset of the UR partition from the TRS into predeterminedly associated RRSs (FIGS. 9, 15, 16 for commit and FIGS. 10, 13, 14 for abort); and

(d) at the termination (END) of the UR state changes utilizing the RRSs, using said log manager to discard all log data identified on the hard log as belonging to the given UR partition from the TRS rendering space occupied by said discarded log data on said hard log amenable to reuse (FIGS. 11, 17-24).