Automates the configuring/running of a quick and dirty, self contained, replicated and sharded MongoDB environment on a local Linux host machine, to help a user explore these features - not to be used for real production environments.
By default, creates an environment containing the following 11 server processes, listening on different local ports:
--REPLICA-0 mongod
SHARD0--REPLICA-1 mongod
--REPLICA-2 mongod
--REPLICA-0 mongod
SHARD1--REPLICA-1 mongod
--REPLICA-2 mongod
--CONFIG-SVR-1 mongod
CONFIG--CONFIG-SVR-2 mongod
--CONFIG-SVR-3 mongod
--MONGOS-SVR-0 mongos
ROUTER
--MONGOS-SVR-1 mongos
- Tested on Linux only (Ubuntu 14.04 & 16.04 x86-64 - other Linux variants should be fine). Mac OS X hosts may be ok too, but haven't tested.
- Requires recent version on MongoDB (eg. 3.2.x, 3.4.x or 3.6) already installed on the local machine - versions earlier than 3.2 are not supported due to the project's use of the preferred modern way to configure a "config db" as a "Config Server Replica Set", which is not supported in earlier versions.
-
In this root directory, modify file 'localenv' to reflect the local settings required
-
From the terminal, run the following to delete and re-create the directories to hold Mongo server data files and logs (also kills any mongod/mongos processes that happen to be running on the local machine):
$ ./1_rebuild_target_directories.sh -
From the terminal, run the following to configure and start the servers with replication and sharding (this takes a few minutes when first run, if not using "smallfiles", whilst mongod server initialises its data files)
$ ./2_start_servers.sh -
OPTIONAL From the terminal, run the following to insert lots of small randomly generated documents into the sharded collection, using a small range of values for the shard key
$ ./3_inject_docs.sh
You should now be able to connect to one of the mongos processes, using the Mongo Shell, and run sh.status() to view the status of the cluster.