Managing checkpoint queues in a multiple node system

• US 7,296,039 B2
• Filed: 10/12/2004
• Issued: 11/13/2007
• Est. Priority Date: 11/24/1998
• Status: Expired due to Term

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Chan, Wilson Wai Shun, Srihari, Vinay, Brower, David, Bridge, William H., Bamford, Roger J., Chandrasekaran, Sashikanth, MacNaughton, Neil
• Primary Examiner(s): Corrielus; Jean M.
• Application Number:
  US10/965,360
• Publication Number:
  US 20050065907A1
• Time in Patent Office: 1,127 Days
• Field of Search: 707/100, 707/202, 707/201, 707/200, 707/8, 709/201, 717/108, 717/124
• US Class Current: 711/119
• CPC Class Codes: G06F 11/1469 Backup restoration techniques
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery
• CPC Class Codes: G06F 12/0804 with main memory updating G06F12/0806 takes precedence
• CPC Class Codes: G06F 12/0815 Cache consistency protocols
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users
• CPC Class Codes: Y10S 707/99953 Recoverability
• CPC Class Codes: Y10S 707/99954 Version management
• CPC Class Codes: Y10S 707/99955 Archiving or backup

Abstract

Techniques are provided for managing caches in a system with multiple caches that may contain different copies of the same data item. Specifically, techniques are provided for coordinating the write-to-disk operations performed on such data items to ensure that older versions of the data item are not written over newer versions, and to reduce the amount of processing required to recover after a failure. Various approaches are provided in which a master is used to coordinate with the multiple caches to cause a data item to be written to persistent storage. Techniques are also provided for managing checkpoints associated with the caches, where the checkpoints are used to determine the position at which to begin processing recovery logs in the event of a failure.

Claims

1. A method for writing data items to persistent storage, the method comprising the steps of:

maintaining, within a first node, a first queue of entries that correspond to dirty data items that need to be written to persistent storage;

maintaining, within the first node, a second queue of entries that correspond to dirty data items that need to be written to persistent storage;

in response to determining that a dirty data item, corresponding to a particular entry, needs to be transferred to a node other than said first node, moving said particular entry from said first queue to said second queue;

selecting, at the first node, which data items to write to persistent storage based on which queue, of the first queue and the second queue, an entry for a data item resides in, wherein said selecting gives priority to data items that correspond to entries in said second queue; and

writing, to persistent storage, a particular data item selected based on which queue, of the first queue and the second queue, an entry for the particular data item resides in.

2. The method of claim 1 wherein the step of moving entries includes moving an entry from said first queue to said second queue in response to a message received by said first node, wherein said message indicates that another node has requested the data item that corresponds to said entry.

3. A method for writing data items to persistent storage, the method comprising the steps of:

maintaining a forced-write count for each of said data items;

incrementing the forced-write count of a data item whenever the data item is written to persistent storage by one node for transfer of the data item to another node;

selecting which dirty data items to write to persistent storage based on the write counts associated with the data items; and

writing one or more selected dirty data items, which were selected based on the write counts associated with the selected dirty data items, to persistent storage.

4. The method of claim 3 further comprising the steps of:

storing dirty data items that have forced-write counts above a certain threshold in a particular queue; and

when selecting dirty data items to write to persistent storage, giving priority to data items stored in said particular queue.

5. A computer-readable medium carrying instructions for writing data items to persistent storage, the instructions, which when executed by one or more processors, causes:

maintaining, within a first node, a first queue of entries that correspond to dirty data items that need to be written to persistent storage;

maintaining, within the first node, a second queue of entries that correspond to dirty data items that need to be written to persistent storage;

in response to determining that a dirty data item, corresponding to a particular entry, needs to be transferred to a node other than said first node, moving said particular entry from said first queue to said second queue;

selecting, at the first node, which data items to write to persistent storage based on which queue, of the first queue and the second queue, an entry for a data item resides in, wherein said selecting gives priority to data items that correspond to entries in said second queue; and

writing, to persistent storage, a particular data item selected based on which queue, of the first queue and the second queue, an entry for the particular data item resides in.

6. The computer-readable medium of claim 5 wherein the step of moving entries includes moving an entry from said first queue to said second queue in response to a message received by said first node, wherein said message indicates that another node has requested the data item that corresponds to said entry.

7. A computer-readable medium carrying instructions for writing data items to persistent storage, the instructions, which when executed by one or more processors, causes:

maintaining a forced-write count for each of said data items;

incrementing the forced-write count of a data item whenever the data item is written to persistent storage by one node for transfer of the data item to another node;

selecting which dirty data items to write to persistent storage based on the write counts associated with the data items; and

writing one or more selected dirty data items, which were selected based on the write counts associated with the selected dirty data items, to persistent storage.

8. The computer-readable medium of claim 7 further comprising instructions for performing the steps of:

storing dirty data items that have forced-write counts above a certain threshold in a particular queue; and

when selecting dirty data items to write to persistent storage, giving priority to data items stored in said particular queue.