Getting started: Using data containers

The state of a Spawn data container can easily be saved so that you can

• undo mistakes
• return to previous states
• store the state to make further copies from

note

Ensure you've completed the Spawn installation guide before going further.

Accidentally delete data from the database

We're going to pretend to make a mistake with the database, deleting some data that we didn't intend to. Using the connection details, connect to the database using your preferred client and delete all the records from a table:

PostgreSQL

DELETE FROM public.film_actor;

MySQL

DELETE FROM sakila.film_actor where actor_id > 0;

SQL Server

DELETE FROM Sales.InvoiceLines;

Reset the database to its original state

Normally we'd now need to restore the data from a backup, but Spawn allows you return the database to its original state in just a few seconds using the reset data-container command:

spawnctl reset data-container spawn-tutorial

Reconnect to the database and check -- the data we just deleted is restored, undoing the mistake.

Make some changes you'd like to keep

What if you want to keep changes that you've made, so that a reset will only return to that point? You can save the state of your database at any time using the save data-container command.

Let's add a new table and some data to our database. Use your database client to execute the following:

PostgreSQL

CREATE TABLE public.crew

(

crew_id integer NOT NULL,

first_name text COLLATE pg_catalog."default" NOT NULL,

last_name text COLLATE pg_catalog."default" NOT NULL,

CONSTRAINT crew_pkey PRIMARY KEY (crew_id)

);

INSERT INTO public.crew VALUES (1, 'John', 'Smith');

INSERT INTO public.crew VALUES (2, 'Jane', 'Doe');

MySQL

CREATE TABLE sakila.crew

(

crew_id integer NOT NULL,

first_name text NOT NULL,

last_name text NOT NULL,

CONSTRAINT crew_pkey PRIMARY KEY (crew_id)

);

INSERT INTO sakila.crew VALUES (1, 'John', 'Smith');

INSERT INTO sakila.crew VALUES (2, 'Jane', 'Doe');

SQL Server

CREATE TABLE [Warehouse].[StockSection]

(

[StockSectionID] [bigint] IDENTITY(1,1) NOT NULL,

[StockSectionName] [nvarchar](100) NOT NULL,

CONSTRAINT [PK_Warehouse_StockSection] PRIMARY KEY NONCLUSTERED ([StockSectionID] ASC)

)

GO

INSERT INTO [Warehouse].[StockSection] VALUES ('East A1');

INSERT INTO [Warehouse].[StockSection] VALUES ('East A2');

Save the new state as a "revision"

We can save the current state of the database as a new revision. This is the default state that reset will return to. Each time you use the save data-container command it will create a new revision.

spawnctl save data-container spawn-tutorial

Successfully saved data container 'spawn-tutorial'

New revision is 'rev.1'

note

You can use the load data-container command to return to states other than the current revision.

Store the state to make further copies

The revisions that we've created only exist for as long as the data container lives -- if we delete the data container the revisions will be lost. To store a revision for later use we need to use the graduate data-container command, creating a persistent data image that can be used to create new data containers.

spawnctl graduate data-container spawn-tutorial --revision rev.1 --name spawn-tutorial-new-table

Successfully graduated data container 'spawn-tutorial' at revision 'rev.1' to a new data-image

New image 'spawn-tutorial-new-table' (97480) available!

The new image appears in the list of available data images:

spawnctl get data-images

This data image includes the new table that we added to our database, and can be used to create new data containers whenever we need them:

spawnctl create data-container --image spawn-tutorial-new-table --lifetime 1h