Ions Reference

Ion applications are described by a standard Clojure deps.edn file. Add:

• com.datomic/ion as a runtime dependency
• com.datomic/ion-dev under a :dev alias
• a "datomic-cloud" entry under your :mvn/repos

;; add the following to get the latest ion release
{:deps {com.datomic/ion {:mvn/version "0.9.7"}}
 :mvn/repos {"datomic-cloud" {:url "s3://datomic-releases-1fc2183a/maven/releases"}}
 :aliases
  {:dev {:extra-deps {com.datomic/ion-dev {:mvn/version "0.9.156"}}}}}

See ion-starter's deps.edn for an example.

An ion function is a plain function, with a signature based on its intended use:

To write an ion, implement a function with one of these signatures. A single ion application can have any number of ions of any types.

The datomic.ion.starter namespace has examples of each type of ion signature.

An ion application must have a resource named datomic/ion-config.edn on its classpath. This resource tells Datomic how to invoke your application:

• the :app-name key names the CodeDeploy application. If you did not override this name when creating your Datomic compute group, it will default to your system name.
• the :allow key whitelists ions that Datomic is allowed to call. When you deploy an application, Datomic will automatically require all the namespaces mentioned in the :allow section.
• the :lambdas section configures lambda entry points for your application.

The example below is taken from the ion-starter project.

{:allow [;; transaction function
         datomic.ion.starter/inc-attr

         ;; query function
         datomic.ion.starter/modes

         ;; lambda handler
         datomic.ion.starter/echo

         ;; web application
         datomic.ion.starter/get-tutorial-schema    
         ]            
 :lambdas {:echo
           {:fn datomic.ion.starter/echo
            :description "Echos input"}
           :get-tutorial-schema
           {:fn datomic.ion.starter/get-tutorial-schema
            :description "returns the schema for the Datomic docs tutorial"}}
 :app-name "<YOUR-APP-HERE>"}

When you are developing ions locally, you will want to connect to a Datomic cloud system as a client. But when you deploy ions, you will want the same code to utilize an in-memory implement of client. The :ion server-type implements this behavior, connecting as a client when remote, and providing an in-memory implemention of client when running in Datomic Cloud.

The following example shows how to create a client for an ion application:

(require '[datomic.client.api :as d])

(def cfg {:server-type :ion
          :region "<your AWS Region>" ;; e.g. us-east-1
          :system "<system-name>"
          :query-group "<system-name>"
          :endpoint "http://entry.<system-name>.<region>.datomic.net:8182/"
          :proxy-port <local-port for SSH tunnel to bastion>})

(def client (d/client cfg))

The datomic.ion.dev main function takes the following arguments for push:

You can specify a Named Profile by passing it as :creds-profile. You can override your environment's default region by providing :region.

As a development convenience, Ions support an unreproducible push, allowing you to deploy work-in-progress code not committed to git. Since there is no git SHA, you must specify an unrepro name (:uname) when pushing. You are responsible for making the uname unique within your org+region+app.

It is ok to push to the same :uname repeatedly, so long as you understand that the last push wins.

Push will ensure that git and maven libs exist under

s3://$(datomic-code-bucket)/datomic/libs

and will create a CodeDeploy application revision located at

s3://$(datomic-code-bucket)/datomic/apps/$(app-name)/$(git-sha).zip

or at

s3://$(datomic-code-bucket)/datomic/apps/$(app-name)/$(uname).zip

if a :uname is specified.

The datomic.ion.dev main function takes the following arguments for deploy:

The :group key takes compute group name as a value. If you followed the upgrade instructions and are running separate storage and compute stacks the name of your compute group is the name you gave the compute stack when you created it.

However, if you launched from AWS Marketplace, you are running a nested stack with an automatically generated compute stack name. To find the name of your compute stack, go to the AWS CloudFormation console and look through the list of Stack Names. The name of your compute stack will be $(SystemName)-Compute-$(GeneratedId). Use that whole stack name as the parameter to :group.

You can specify a Named Profile by passing it as :creds-profile. You can override your environment's default region by providing :region.

When you deploy an ion application, Datomic will use an AWS Step Machine to

• CodeDeploy the code for the application onto each instance in the compute group. This deployment is one-at-a-time, so the group will remain available with N-1 members active throughout the deployment.
• Ensure that all the lambdas requested via the ion-config.edn map exist and are configured correctly.

Deploy runs a step function that deploys the application to CodeDeploy and ensures that the lambdas are correctly configured.

You can specify a Named Profile by passing it as :creds-profile. You can override your environment's default region by providing :region.

Deploy Status returns the status of the Code Deploy and the ion deploy as a map:

{:deploy-status "SUCCEEDED"
 :code-deploy-status "SUCCEEDED"}

See the Step Functions reference for possible values of the status keys.

You can monitor the deployment in near real-time from the Step Functions Console. Look for a state machine named "datomic-$(group)".

A web ion is a function that takes the following input map:

and returns

where the :body can be an InputStream, String, or File.

The input and output maps are mostly compatible with the Clojure Ring Spec, and many web applications should be able to run unmodified as web ions.

The ion-starter project includes an example web ion.

Web services are exposed as lambda proxies through the AWS API Gateway. The ion library includes a helper function datomic.ion.lambda.api-gateway/ionize that will convert a web ion into a lambda ion, as demonstrated in the ion-starter project:

(def get-tutorial-schema
  "API Gateway web service ion for tutorial-schema-handler."
  (apigw/ionize tutorial-schema-handler))

The lambda ion is then configured like any other lambda. You can then use the API Gateway to:

• deploy an API to the internet
• manage content encodings and compression
• validate requests
• configure access control
• issue and manage API keys

The ions tutorial includes an example that deploys a web ion to the internet.