README
Weaver
Migrate related document objects within and across different databases.
Context
Often times when working with NRDBs like MongoDB, documents reference to other documents from different collections (or even DBs). This becomes a challenge when trying to replicate an interwined dataset to be used in a different environment.
For example, a database has the following collections/documents:
// db.users.findOne({_id: ObjectId('abcdef78901234abcdef1234')})
{
_id: ObjectId('abcdef78901234abcdef1234'),
name: 'John',
orders: [ { orderId: '4321fedcbafedcba67890123' } ]
}
// db.orders.findOne({_id: ObjectId('4321fedcbafedcba67890123')})
{
_id: ObjectId('4321fedcbafedcba67890123'),
cartId: 'fedcba67890123fedcba4321'
}
// db.carts.findOne({_id: ObjectId('fedcba67890123fedcba4321')})
{
_id: ObjectId('fedcba67890123fedcba4321'),
userId: 'abcdef78901234abcdef1234'
}
Note how user
relates to order
, and order
relates to cart
.
If you wanted to migrate the orders
and carts
associated to the user
to your local environment, you'd need to:
- Go to the
[users|orders|carts] collection
and find the document. - Check if any of the fields is a reference to another collection.
- Copy the reference value.
- Repeat from step 1.
This tool finds the relationships, and migrates them to a destination db:
Query | source db | target db |
---|---|---|
db.users.findOne(ObjectId('abcdef78901234abcdef1234')) |
1 | 1 |
db.orders.findOne(ObjectId('4321fedcbafedcba67890123')) |
1 | 1 |
db.carts.findOne(ObjectId('fedcba67890123fedcba4321')) |
1 | 1 |
It can also plot the entity-relationship graph (coming soon).
Usage
Requirements
Getting started
Weaver runs on NodeJS >= 10
Install via npm
npm i -g @skelogh/weaver
How to test the CLI in dev mode
Install the CLI tool
- After cloning the repo and installing dependencies, run
npm run build
. - Run
npm pack
on the project root. - A file
weaver-<VERSION>.tgz
will be created. - Run
npm install -g weaver-<VERSION>.tgz
, it will install the package globally.
Using the CLI tool
- After installing, just issue the
weaver
command, the CLI help will display the CLI help:
Usage: weaver [OPTIONS] COMMAND [ARG...]
weaver [ --help | -v | --version ]
Commands:
weaver run Runs the app with the loaded configuration
weaver add Creation of client, query or ignore
weaver remove Removal of clients, queries or ignores
Options:
--version, -v Print version information and quit
--config, -c Read or set path of config file, default: undefined
--dry Run but don't save
--info Displays the current settings
--json Write the output in the configured JSON file
--json-file The JSON filepath where output will be streamed to
--limit The max amount of docs to retrieve
--queries, --qq Document ids to get relationships from, e.g.: 2a3b4c5d6e7f8g9h2a3b4c5d e7f8g9h2a3b4c5d2a3b4c5d6
--verbose, -V Enable highest level of logging, same as DEBUG=*
--help Show help
1. Add clients
In the CLI, run weaver add client
, the following help will display:
Usage: weaver add client -fntou --options.<option_name>
-f [mongodb] <String> The client db family, mongodb is only supported for now.
-n <name> <String> The client db name.
-t [source|target] <String> source if the data will be pulled from it, target otherwise.
-o [<source.name>] <String> The target's origin db where the data will be copied from.
-u <url> <String> The client db URL.
--options.<opt_name> <Any> Client db-specific options, for now MongoClient options, use dot notation to set each option
Example: --options.readPreference secondaryPreferred
To pull data from a db into another, you'll need to add both the source
and target
clients. The source
is where the document we want is stored, and target
is where you want to copy to.
Example:
Assume the following remote server settings for source
:
- Family: mongodb
- Type: source
- IP: 172.17.0.4
- Port: 27017
- Database name: production
weaver add client -f mongodb -n production -t source -u mongodb://172.17.0.4:27017
Assume the following local server settings for target
:
- Family: mongodb
- Type: target
- IP: 127.0.0.1
- Port: 27017
- Database name: development
- Origin: production (This is used to map where the data will be pulled from, useful when dealing with multiple
source
andtarget
databases.)
weaver add client -f mongodb -n development -t target -o production -u mongodb://127.0.0.1:27017
2. Find the document to clone
From the source
database, find the document you want to copy to the target
database and copy the stringified _id
:
// Assume the following mongodb document in the `production` database, `pets` collection:
{
_id: ObjectId("507f1f77bcf86cd799439011"),
name: "fluffy"
}
3. Clone the document(s)
Once steps 1 & 2 are done, just run the following command:
weaver run --queries 507f1f77bcf86cd799439011
4. See the results
If all is setup correctly, you should now have a copy of the document in your target
database, it will create the collection even if it didn't exist before:
> mongodb
mongodb> use development
mongodb> db.pets.find(ObjectId("507f1f77bcf86cd799439011"))
{
_id: ObjectId("507f1f77bcf86cd799439011"),
name: "fluffy"
}
>
Roadmap
- [WIP]
weaver
CLI (follow the issue) - Add plugins for different data sources.
MongoDB- ElasticSearch
- DinamoDB
- Cassandra
- and so on...
- API/SDK (?)
- Client UI
- Connect via SSH.
- Doc pages
Questions? create an issue!.