Skip to main content

MongoDB Backup and Restore

It's a general need in MongoDB development, to take backups or restore DB with an old backup. DB backup can be done at both DB level and individual Collection level. Let's see how we can perform different backup/restore operations..

DB Backup

Complete DB

Assuming database to backup is "mydb", the best thing is to back it as gzip to save some space on your Server.
mongodump --archive=mydb.2017-09-21.gz --gzip --db mydb
This will create the archive in the directory where you're executing the command.

Another way is to take backup as bson documents.
mongodump -h localhost -p 27017 -d mydb -o C:\mongobackup\20170921
A directory with db name will be created under C:\mongobackup\20170921
In case you have multiple instances running on same server, you can use --host and --port options.

Individual Collection

mongodump --collection myCollection --db mydb
This will create backup with same name as that of collection (myCollection in this case).

Restore Database

Restoring from an archive

mongorestore --gzip --archive=mydb.2017-09-21.gz --db mydb

Restoring from bson

Assuming you took backup in C:\mongobackup\20170921
mongodump -h localhost -p 27017 -db mydb C:\mongobackup\20170921\mydb

Comments

Post a Comment

Popular posts from this blog

MongoDB BulkWrite Java API

Since version 3.2, MongoDB has introduced Bulk Update methods. In context of RDBMS, it's like SQL Batch Jobs, where SQL Statements are prepared in different chunks and a batch of statements are submitted to DB for update/insert. Here are some important points about MongoDB Bulk Write operation.. Useful in case you've huge data to update/insert. Mongo automatically prepares batches (of 1000 default) and start execution in an ordered/unordered manner. This drastically reduce DB trip time. Let's say there are 50 thousand records to update, now instead of 50k round trips to DB from your app server, using Bulk Update it would be reduced to just 50 round trips. Let's see an example below: List<WriteModel<Document>> updateDocuments = new ArrayList<WriteModel<Document>>(); for ( Long entityId : entityIDs ) { //Finder doc Document filterDocument = new Document (); filterDocument . append ( "_id" , ent

MongoDB Aggregation using Java API

A very common problem scenario in programming is to get the records or record count by certain fields. For developers familiar with RDBMS, it's like creating a SQL with combination of count function and group by attributes. For MongoDB too, it's very similar. Let's look at the example below fetching no of employees group by department Ids. public Map < Long , Integer > getEmployeeCountMapByDeptId () { Map < Long , Integer > empCountMap = new HashMap <>(); AggregateIterable < Document > iterable = getMongoCollection (). aggregate ( Arrays . asList ( new Document ( "$match" , new Document ( "active" , Boolean . TRUE ) . append ( "region" , "India" )), new Document ( "$group" , new Document ( "_id" , "$" + "deptId" ). append ( "count"

MongoDB fetch operation using Java API

With changes in MongoDB, there has been several changes in its Java API as well. There are now some good and easy way to perform different operation with DB. MongoDB Java API is really very simple and easy to understand. If we understand the basics, we can build up simple to complex queries. Let's take a look at the example of MongoDB fetch operation using the new Java API and then we'll try to understand the basics.. public MyEntity findMyEntityById ( long entityId ) { List < Bson > queryFilters = new ArrayList <>(); queryFilters . add ( Filters . eq ( "_id" , entityId )); Bson searchFilter = Filters . and ( queryFilters ); //Fields to return. //_id is available by default. A value of "0" would skip it List < Bson > returnFilters = new ArrayList <>(); returnFilters . add ( Filters . eq ( "name" , 1 )); returnFilters . add ( Filters . eq ( "_id" , 0 ))