chat

Chat

An example chat application.

%%{init: {"flowchart": {"defaultRenderer": "elk"}} }%%
graph TD
    %% Nodes.
    github.com/ServiceWeaver/weaver/Main(weaver.Main)
    github.com/ServiceWeaver/weaver/examples/chat/ImageScaler(chat.ImageScaler)
    github.com/ServiceWeaver/weaver/examples/chat/LocalCache(chat.LocalCache)
    github.com/ServiceWeaver/weaver/examples/chat/SQLStore(chat.SQLStore)

    %% Edges.
    github.com/ServiceWeaver/weaver/Main --> github.com/ServiceWeaver/weaver/examples/chat/ImageScaler
    github.com/ServiceWeaver/weaver/Main --> github.com/ServiceWeaver/weaver/examples/chat/LocalCache
    github.com/ServiceWeaver/weaver/Main --> github.com/ServiceWeaver/weaver/examples/chat/SQLStore

How to Run Locally

First, run a local MySQL instance and initialize it with chat.sql. We recommend using Docker for this:

# Run the MySQL instance.
$ docker run \
      --rm \
      --detach \
      --name mysql \
      --env MYSQL_ROOT_PASSWORD="password" \
      --env MYSQL_DATABASE="chat" \
      --volume="$(realpath chat.sql):/app/chat.sql" \
      --publish 127.0.0.1:3306:3306 \
      mysql

# Wait about ten seconds for the MySQL instance to start. Then, initialize the
# chat database with chat.sql.
$ docker exec mysql sh -c "mysql --password=password < /app/chat.sql"

Next, update the db_uri field in weaver.toml to point to your MySQL instance. If you used the Docker commands above, the default value of db_uri should already point to your database. You don't have to change anything.

Finally, run the application.

$ go build .

# Run the application in a single process.
$ weaver single deploy weaver.toml

# Run the application in multiple processes.
$ weaver multi deploy weaver.toml

How to Run on Kubernetes

First, run a MySQL instance in your Kubernetes cluster and initialize it with chat.sql. We recommend using mysql.yaml for this:

$ kubectl apply -f mysql.yaml

Next, update the db_uri field in weaver.toml to point to your MySQL instance. If you're using mysql.yaml, then you should update the SQLStore section of weaver.toml to have the following contents:

["github.com/ServiceWeaver/weaver/examples/chat/SQLStore"]
db_driver = "mysql"
db_uri = "root:password@tcp(mysql:3306)/chat"

Then, update the repo field in weaver.toml with the Docker repository where the chat application will be uploaded. For example, if your Docker Hub username is myusername, you can update the [kube] section of weaver.toml to the following:

[kube]
listeners.chat = {public = true}
repo = "docker.io/myusername/"

See weaver kube deploy --help for more information on the repo field.

Finally, deploy the application using weaver kube.

$ go build .
$ kubectl apply -f $(weaver kube deploy weaver.toml)

How to Run on GKE

Create a new MySQL instance on Cloud SQL. During creation, ensure the "No password" and "Private IP" options are checked and the latter is configured to use the "default" network.

Once the instance is created, copy the IP address from the instance dashboard and substitute it for localhost in the weaver.toml file.

Deploy the app:

$ weaver-gke deploy ./weaver.toml