Current state of MongoDB Backup options
summary
Covers practical backup behavior and tradeoffs: mongodump goes to secondary via --host but mongos hits primary so logical dumps are “squishy,” restores historically take 3–10 days on large restores, LVM snapshots need fsync/lock or risk partial data, and NFS/EBS have caveats with mmapv1. Mentions Ops Manager $2/GB and its oplog-stream plus snapshot metadata model. Describes Percona Labs’ consistent-backup (mcb1 promoted to production) that spiders shards, tails oplogs until all shard dumps finish, uploads to S3, and currently supports compression, YAML config, Docker and RPM packaging.
AI-written summary · about
chapters (9)
- 0:00MongoDB backup overview
- 3:07Backup types explained
- 6:17LVM snapshots & storage
- 10:12Hot backups & enterprise tools
- 13:02Sharding backup issues
- 16:12Fixing sharded backups
- 17:17Consistent Backup Tool
- 22:35Future backup tooling
- 25:11Closing & Q&A
related talks
40mMongoDB Backups Overview
39mPhysical Backups with Percona Backup for MongoDB
26mConsistent Backups for Multi-Shard MongoDB
52mPBM - The Backup Open Source Tool for MongoDB
07mPBM -The backup OpenSource tool for MongoDB
09mOpen Software for MongoDB Where Percona Fits
49mCreating Backups for Various Transactional DBMSs Under Load
27mBackup and Restore: Discover Best Practices for a Distributed Database Backup Strategy