Datomic Ion Parameters look good. All makes sense and I'm glad to see it added. +1

One thing which got my attention was the example. It shows using it to discover the db-name. That struck me as unnecessary since there can only be one ion app per datomic ion stack. Can someone give me an example of why it would be useful to have a configurable db name?

Hello, I'm trying to connect the Datomic console to an in memory database (local dev). Is it feasible?

Found this from 3 yrs ago, says “Console cannot connect to in-memory” http://datomic.narkive.com/4RyKpVMd/can-the-console-be-used-against-either-an-in-memory-or-dev-database

There’s no such thing as read-only Datomic, right? Like if App A wants to look at App B’s data without ever modifying it (e.g. a reporting front end), I can’t just give App A read-only access to App B’s data.

not that I know of, your app must enforce this

Thanks @manutter51. That is sad since I have a very smooth dev workflow with integrant. I can reset the whole database with fresh dev data after a simple cider-refresh. Also I do everything in the same JVM instance. I wonder if there is a similar dev workflow that makes it possible to also easily have the datomic console.

@dam https://docs.datomic.com/on-prem/dev-setup.html

you’d need to use dev storage

not mem

but you gain console and disk-based persistance

Thanks @marshall. I saw that do but it doesn't seem easy to reset the database from the REPL. I made a mistake in my last post. I don't use cider-refresh but a custom reload-with-dev-data from the dev namespace with integrant.

(defn reload-with-dev-data []
  (halt)
  (go)
  (dev-data/load-dev-data! @db-conn-ref))

sure; call the same functions

@dam for dev purposes you might also want to check out datomock: https://github.com/vvvvalvalval/datomock

Thanks @timgilbert, I'll check that. Also I see Valentin often during Paris Clojure Meetups. I'll discuss about that with him directly next time (ping @val_waeselynck (Salut Val !)).

hi. have a datomic cloud question for any of the cognitect folks who might be around. We have a pretty complex VPC setup, and one of our requirements is that traffic from any of our ‘internal’ VPCs (dev,test,etc and associated datomics) has to traverse a transit VPC to get out to the world. It looks like the IGW’s that get created are used for license checking or something? We tried just disabling one and things just broke 🙂 . In any case, we wanted to get some input from you guys on how we might best accomodate this..

@eoliphant Datomic uses the internet gateway to access some AWS services

I don’t recall which ones off the top of my head, but any AWS service for which there are not VPC endpoints available or which we haven’t yet been able to transition to using VPC endpoints

ah ok

let me let my guys know. we’ll try to figure it out but if you guys could scare up a list that’d be great

everything but ddb and s3

i suspect Cloudwatch, possibly codedeploy

maybe IAM

SNS / SFN @marshall?

probably so

EFS probably is internal traffic only... administrative endpoints notwithstanding

yeah, i think EFS is strictly scoped to VPC

lambda for ions uses SNS

lambda service itself may also be a public endpoint

yeah i was wondering about the lambdas. and yeah I think my dev noticed the issue during a deploy. SO codedeploy is a likely place to look

deploy will definitely require lambda and step functions

with datomic cloud supporting :limit and :offset is there planned support for :sort-by perhaps?

nevermind apparently the sort aggregate is a thing (but I can not see it in the docs?)