Percona Backup for MongoDB 2.6.0 (2024-09-02)¶

Installation

Percona Backup for MongoDB is a distributed, low-impact solution for creating consistent backups of MongoDB sharded clusters and replica sets, and for restoring those backups to a specific point in time.

Release Highlights¶

This release provides the following features and improvements:

Multiple storages for backups and restores¶

You can now define multiple storages for making backups and instruct PBM on what storage to save a backup: on the main or on any additional (external) one. This ability helps you save on data transfer costs when using cloud storage, as well as enables you to follow closely with the requirements of your organization’s backup policy.

To learn more, read Multiple storages for backups.

Adjust node priority for point-in-time recovery oplog slicing¶

You can now redefine node priorities not only for making backups but also for point-in-time recovery oplog slices. For example, you can instruct PBM to save oplog slices either from a dedicated node or from nodes in the geographically closest location, thus reducing network latency.

See the Adjust node priority for oplog slices section for guidelines.

Additional control over PBM command execution¶

You can now configure the waiting time for a command execution by passing the --wait-time flag for the following commands:

pbm config –force-resync –wait –wait-time=”5m”
pbm backup –wait –wait-time=”5m”
pbm restore –wait –wait-time=”5m”
pbm oplog-replay –wait –wait-time=”5m”
pbm profile add –sync –wait –wait-time=”5m”
pbm profile remove –wait –wait-time=”5m”
pbm profile sync –wait –wait-time=”5m”
pbm delete-pitr –older-than –wait –wait-time=”5m”
pbm cleanup –older-than –wait –wait-time=”5m”

This way you have more control over the PBM operation. This enhancement also improves the error logging.

Snapshot-based physical backups are generally available¶

With snapshot-based physical backups now generally available, you can use this functionality in production environments and enjoy all the benefits of faster restores from snapshots with almost immediate access to data.

Dropped support of MongoDB 4.4¶

Percona Backup for MongoDB drops support of MongoDB 4.4. Existing functionality in Percona Backup for MongoDB remains compatible with MongoDB 4.4 and Percona Server for MongoDB 4.4; however, further enhancements and bug fixes are no longer tested against this version.

Packaging Changes¶

Percona Backup for MongoDB is available on Ubuntu 24.04 (Noble Numbat).
Percona Backup for MongoDB is no longer supported for Ubuntu 18.04 (Bionic Beaver), Debian 10 and Red Hat Enterprise 7 and derivatives as these operating systems reached End-Of-Life.

Join Percona Squad¶

Participate in monthly SWAG raffles, get an early access to new product features and invite-only “ask me anything” sessions with database performance experts. Interested? Fill in the form at squad.percona.com/mongodb.

Changelog¶

New Features¶

PBM-1228 - Support multiple storage for backups and restores.

Improvements¶

PBM-341 - Fixed the S3 storage URL in text output when starting a backup.
PBM-901, PBM-1323 - Show node priorities in pbm status output
PBM-920, PBM-940, PBM-1043 - Take a node priority into account for oplog slicing
PBM-1303 - Add the ability to configure the waiting time for commands execution
PBM-1320 - Improve logging for a logical restore progress if a node to make a restore is not the primary one
PBM-1329 - Improve saving backups to a filesystem storage by writing to the temporary files before renaming to the final filename.
PBM-1376 - Add support for SSE-S3 for S3 providers

Bugs Fixed¶

PBM-557 - Restart the election process and elect a new node to do a PITR oplog slicing if the initial node became primary
PBM-1190 - Remove excessive parameter from the MongoDB connection URI string in the environment file for the pbm-agent
PBM-1208 - Fixed the issue with hanging backups after pbm-agent is restarted by adding an active lock check before running the backup
PBM-1318 - Restore printing the uploadPartSize and maxUploadParts values in the log
PBM-1319 - Fixed the issue with inability to make a backup from the primary node even with the highest priority assigned
PBM-1326 - Resume PITR oplog slicing when recovering from a critical failure during a backup. (Thanks Daniel Oliver for reporting the issue)
PBM-1332 - Fixed the issue with physical backup not being completely removed from the storage after executing pbm delete-backup or pbm cleanup commands.
PBM-1341 - Fixed the issue with failing physical restore on a node shutdown stage by retrying the shutdown operation.
PBM-1343 - Fixed the issue with pbm-agent not being able to start oplog slicing by releasing the lock before exiting the function to allow pbm-agent on another node to continue the slicing. (Thanks to Tomasz Spyrka for reporting the issue and contribution)
PBM-1344 - Continue PITR oplog slicing from a fresh full backup if PBM cannot catch up with old oplog slices on the storage.
PBM-1350 - Hide password from log output
PBM-1360 - Drop the pbm.Lock collection during physical restore to prevent restoring excessive locks
PBM-1364 - Ensure PITR routine is fully operational after the cluster leader restart
PBM-1367 - Disallow making backups of deployments with sharded time series collections
PBM-1239 - Fixed the issue with logical backups failing on GCS by using the libraries that support retries
PBM-1348 - Fixed the incorrect behavior of the shell when using the --wait for the JSON output in pbm commands
PBM-1377 - Fixed the issue with backups failing to start when there is insufficient oplog range for PITR
PBM-1057 - Prevented PBM to start backup and oplog slicing on nodes that do initial sync

Get expert help¶

If you need assistance, visit the community forum for comprehensive and free database knowledge, or contact our Percona Database Experts for professional support and services.

Community Forum Get a Percona Expert

Last update: September 13, 2024
Created: September 13, 2024