Optimistic reads in a multi-node environment
View Patent ↗A method, system, computer system, and computer program product to support a distributed environment in which changes to data shared by multiple nodes are logged using private logs and managed by a coordinator. The coordinator recognizes invalidating operations and informs a reader when data need to be read again after an invalidating operation corrupted and/or caused the copy of the data read by the reader to be out-of-date. The coordinator can provide a location from which current valid data can be read. The coordinator can use an operations history to detect invalidating operations and operations trackers to determine when a node is not synchronized with the coordinator.
1. A method comprising:
obtaining a request to perform a read operation on data shared by a plurality of network nodes, wherein
one or more of the network nodes is coupled via a network connection to a shared storage,
the shared storage is configured to store the data,
the read operation is begun substantially simultaneously with sending the request, and the read operation is performed by a reader;
determining whether an invalidating operation to the data has occurred; and
when the invalidating operation has occurred, notifying the reader to re-read the data.
2. The method of claim 1 wherein
the invalidating operation comprises at least one of:
a write operation to the data,
a configuration of storage in which the data are stored,
a configuration of a cluster managing the data,
a partition of the cluster, and
a lock operation on the data.
3. The method of claim 1 , further comprising:
providing a location from which to re-read the data.
4. The method of claim 3 , wherein
the location comprises at least one change to the data.
5. The method of claim 1 , wherein
the determining whether the invalidating operation has occurred comprises determining whether the invalidating operation has occurred within a given timeframe.
6. The method of claim 1 , wherein
the determining whether the invalidating operation has occurred comprises reading a history of at least one invalidating operation.
7. The method of claim 6 , wherein
the at least one invalidating operation occurred within a given timeframe.
8. The method of claim 6 , wherein
the at least one invalidating operation occurred as one of a given number of a plurality of invalidating operations.
9. The method of claim 6 , wherein
the history comprises a respective region of the data affected by each invalidating operation, and the read operation comprises reading at least one respective region.
10. The method of claim 6 , wherein
the history comprises a respective operations tracker value associated with each invalidating operation.
11. The method of claim 10 wherein
the respective operations tracker value associated with each invalidating operation is a counter of invalidating operations on the data.
12. The method of claim 10 wherein
the respective operations tracker value associated with each invalidating operation is a timestamp for the invalidating operation.
13. The method of claim 10 wherein
the notifying the reader to re-read the data comprises providing a current time as a value of one respective operations tracker.
14. A system comprising:
obtaining means for obtaining a request to perform a read operation on data shared by a plurality of network nodes, wherein
one or more of the network nodes is coupled via a network connection to a shared storage,
the shared storage is configured to store the data,
the read operation is begun substantially simultaneously with sending the request, and the read operation is performed by a reader;
determining means for determining whether an invalidating operation to the data has occurred; and
notifying means for notifying the reader to re-read the data when the invalidating operation has occurred.
15. The system of claim 14 further comprising:
providing means for providing a location from which to re-read the data.
16. The system of claim 14 further comprising:
reading means for reading a history of at least one invalidating operation.
17. The system of claim 14 further comprising:
providing means for providing a value of a current time for one respective operations tracker.
18. A computer-readable medium comprising:
obtaining instructions configured to obtain a request to perform a read operation on data shared by a plurality of network nodes, wherein
one or more of the network nodes is coupled via a network connection to a shared storage,
the shared storage is configured to store the data,
the read operation is begun substantially simultaneously with sending the request, and the read operation is performed by a reader;
determining instructions configured to determine whether a invalidating operation to the data has occurred; and
notifying instructions configured to notify the reader to re-read the data when the invalidating operation has occurred.
19. The computer-readable medium of claim 18 further comprising:
providing instructions configured to provide a location from which to re-read the data.
20. The computer-readable medium of claim 18 further comprising:
reading instructions configured to read a history of at least one invalidating operation.
21. The computer-readable medium of claim 18 further comprising:
providing instructions configured to provide a value of a current time for one respective operations tracker.
22. A computer system comprising:
a processor configured to execute instructions;
a memory configured to store the instructions, wherein the instructions comprise:
obtaining instructions configured to obtain a request to perform a read operation on data shared by a plurality of network nodes, wherein
one or more of the network nodes is coupled via a network connection to a shared storage,
the shared storage is configured to store the data,
the read operation is begun substantially simultaneously with sending the request, and
the read operation is performed by a reader;
determining instructions configured to determine whether an invalidating operation to the data has occurred; and
notifying instructions configured to notify the reader to re-read the data when the invalidating operation has occurred.
23. The computer system of claim 22 wherein the instructions further comprise:
providing instructions configured to provide a location from which to re-read the data.
24. The computer system of claim 22 wherein the instructions further comprise:
reading instructions configured to read a history of at least one invalidating operation.
25. The computer system of claim 22 wherein the instructions further comprise:
providing instructions configured to provide a value of a current time for one respective operations tracker.
26. A system comprising:
an obtaining module configured to obtain a request to perform a read operation on data shared by a plurality of network nodes, wherein
one or more of the network nodes is coupled via a network connection to a shared storage,
the shared storage is configured to store the data,
the read operation is begun substantially simultaneously with sending the request, and the read operation is performed by a reader;
a determining module configured to determine whether an invalidating operation to the data has occurred; and
a notifying module configured to notify the reader to re-read the data when the operation has occurred.
27. The system of claim 26 further comprising:
a providing module configured to provide a location from which to re-read the data.
28. The system of claim 26 further comprising:
a reading module configured to read a history of at least one operation.
29. The system of claim 26 further comprising:
a providing module configured to provide a value of a current time for one respective operations tracker.