Staggered Consistent Checkpointing
IEEE Transactions on Parallel and Distributed Systems
Checkpointing for Distributed Databases: Starting from the Basics
IEEE Transactions on Parallel and Distributed Systems
Hi-index | 0.00 |
Checkpointing distributed database systems is useful for recovery from failures as well as for audit purposes. In this paper, we present a non-intrusive checkpointing algorithm for distributed database systems. Our approach uses both checkpoints and transaction logs to capture a transaction-consistent state of the database which helps in reducing overall checkpointing overhead. Out approach is non-intrusive and hence does not block arriving or executing transactions during checkpointing.