Humongous, MongoDB = JSON +Index
Features
document/Object/collection oriented, JSON/BSON, Key/Value, fully indexed(object)
Replication & High Availability, Rich Document based Query and Update
auto-sharding(sharding key) & GridFS, Map/Reduce
ReplicaSet
Basically master/slave replication
automatic fail-over to recovery member nodes
consists of two or more nodes that are copies of each other
automatically elects a primary(master) if there is no primary currently available
Distribute read load
arbiter
Monitoring & Diagnostics
- query profiler
- Http console
- mongostat
- Memory-mapping, keep indexes in memory
- oplog {ts:{},h{}, op:{},ns:{},o:{},o2:{}}, local, db.oplog.
backup
- mongoexport, monogoimport
- mongodump mongorestore
- data consistency use —oplog
- use increamental backup
- read oplog and replay(wordnik tools)
Rollback MongoDB
- Use snapshot + oplog
- Use delayed secondary + oplog
Ops
- Scala up : Vertical
- Scale out: horizontal / auto sharding
- build new index in a huge collections
Scenario
- real time CRUD, web 2.0
- buffer, persistent buffer
- big data, web log
- db cluster
others
no transaction support