Fork me on GitHub

I have a datomic cloud dev env with solo topology I’d like to upgrade to production (to use HTTP direct), can I just delete the old compute stack and create a new one?


@tatut HTTP direct requires a production topology, but you can upgrade the solo to production topology by updating the solo compute stack


note you cannot convert back to solo once you move to production as that is unsupported.


and you can technically delete solo compute and relaunch prod compute against the same storage system, but upgrade is our documented method.


Hi, datomic noob here, I have IONS related question and would appreciate some pointers! I’m on solo topology and would like to expose a ring handler (w/ reitit) through api gateway http api lambda integration. But the router seems to be having trouble with making correct match when accessed through api gateways default endpoint (i.e.

(ns nette.system.web.handler
  (:require [ring.middleware.keyword-params :refer [wrap-keyword-params]]
            [ring.util.response :as resp]
            [reitit.ring :as ring]
            [reitit.core :as r]
            [reitit.ring.middleware.parameters :as parameters]
            [datomic.ion.lambda.api-gateway :as apigw]

(defn router []
   [["/" {:name ::home-page/home-page-view
          :handler  (fn [_]
                      {:status 200
                       :body "Nette: Root URL"})}]]
   {:data {:middleware [parameters/parameters-middleware

(defn app []
     {:path "/"})
     {:not-found (fn [req]
                   {:status 404
                    :body "Nette: Route Not Found"})}))))

(def ion-app
  (apigw/ionize (app)))
Expecting match on “/” w/ response: Nette: Root URL actual: Nette: Route Not Found (default not found handler).


So it seems that something, is getting to the handler, just not what I expect ;)


If I know I have a large amount of Datomic writes incoming, is there a supported way to preemptively scale up DDB?


I imagine you could directly change the read/write throughput on the DynamoDB table. If it’s using autoscaling, you could increase the minimum throughput.