Backup, Restore, and Disaster Recovery
Download as PPTX, PDF1 like682 views
This document discusses backup and disaster recovery strategies for MongoDB databases. It provides an overview of different backup options for MongoDB including mongodump and file system snapshots. It also discusses using replica sets for disaster avoidance by maintaining redundant copies of data across multiple data centers. For sharded clusters, it recommends stopping the balancer, backing up shards and config servers, and resuming the balancer for point-in-time backups. The key messages are that backing up MongoDB is not difficult, there are multiple options to choose from, and practicing backups is important for disaster recovery.
![Disaster Avoidance
> rs.conf()
{
"_id" : "replSetName",
"version" : 3,
"members" : [
{
"_id" : 0,
"host" : "myhost1.dnsname.com:27017“,
“tags” : { “dc” : “east” }
},
{
"_id" : 1,
"host" : "myhost2.dnsname.com:27017“,
“tags” : { “dc” : “east” }
},
{
"_id" : 2,
"host" : "myhost3.dnsname.com:27017“,
“tags” : { “dc” : “west” }
}
]}](https://image.slidesharecdn.com/backups1-130306113416-phpapp01/85/Backup-Restore-and-Disaster-Recovery-16-320.jpg)
Backup, Restore, and Disaster Recovery
- 1. #MongoDBDays Backup, Restore, and Disaster Recovery Craig Wilson Software Engineer, 10gen
- 2. You Gotta Back It Up To Get It Back Leo Laporte
- 4. Slightly Less, But Still Very Important Practice
- 5. Backups with MongoDB • Not difficult – Multiple options – Choose the one that fits your needs • Can be automated • Practice
- 6. Choices
- 7. Choices • Selectivity • Live/Offline • Efficiency • Backup Size • Environment Constraints
- 8. mongodump • Dumps collections to *.bson files – Use bsondump to view these files • Mirrors your structure • Can be run in live or offline mode • --dbpath for direct file access • --oplog to dump oplog last PIT • --query/filter selective dump
- 9. mongorestore • Selected dbs, collections • --oplogReplay
- 10. File System - fsync +lock • No more writes • Copy the data files • Remember to unlock
- 11. File System - Snapshot • Snapshot – Journaling needs to be on – Journal needs to be on the same volume
- 12. File System • Entire database • Backup files will be large • Fastest way to create a backup • Fastest way to restore a backup
- 15. Disaster Avoidance • Redundancy • Multiple Datacenters • Multiple Regions
- 16. Disaster Avoidance > rs.conf() { "_id" : "replSetName", "version" : 3, "members" : [ { "_id" : 0, "host" : "myhost1.dnsname.com:27017“, “tags” : { “dc” : “east” } }, { "_id" : 1, "host" : "myhost2.dnsname.com:27017“, “tags” : { “dc” : “east” } }, { "_id" : 2, "host" : "myhost3.dnsname.com:27017“, “tags” : { “dc” : “west” } } ]}
- 17. Disaster Recovery Logistics • Recovery (restore) time • Replacement – Machine – Component (disk, memory, etc) – Cluster • Configuration (reconfiguration)
- 18. Sharded Systems
- 19. Sharded Systems • No true point in time 1. Stop balancing – sh.stopBalancer() 2. Stop one config server (data R/O) 3. Backup Data (shards, config) 4. Restart config server 5. Resume Balancer
- 20. What have we discussed?
- 21. What have we discussed? • You have Options • It’s not difficult • Practice!
- 22. #MongoDBDays Thank You Craig Wilson Software Engineer, 10gen