pgBackRest 2.01 Released

Posted on 2018-03-28 by Crunchy Data
Related Open Source

March 28, 2018: Crunchy Data announces the release of pgBackRest 2.01, the latest version of the simple, reliable backup and restore system that can seamlessly scale up to the largest databases and workloads.

pgBackRest 2 brings significant improvements to performance. The pgbackrest executable and asynchronous archive-push command are now written in C, which significantly reduces the time required for PostgreSQL to acknowledge WAL segments pushed by the asynchronous process.

pgBackRest supports a robust set of features for managing your backup and recovery infrastructure, including: parallel backup/restore, full/differential/incremental backups, delta restore, parallel asynchronous archiving, per-file checksums, page checksums (when enabled) validated during backup, compression, partial/failed backup resume, backup from standby, tablespace and link support, backup expiration, local/remote operation via SSH, and more.

You can download pgBackRest from the PostgreSQL Yum Repository, the PostgreSQL APT Repository, or from the official pgBackRest website:


New Features

pgbackrest executable is now written in C

pgBackRest commands can now be written in C or embedded Perl. This allows performance-sensitive commands (like asynchronous archive-push) to be implemented in C with significant performance benefits. Most commands are still written in Perl but are loaded only when required to reduce startup time.

Performance Improvements for “archive-push”

The archive-push command is now partially coded in C and allows the PostgreSQL archive_command to run significantly faster when processing status messages from the asynchronous archive process. Testing has shown that the startup time for asynchronous archive-push is up to 40 times faster, enabling more WAL logs to be archived per day.

S3 Performance Improvements

Performance improvements have been made to the S3 list and delete commands.