just thinking of other things that could interfere (or at least maybe could in my mind since I don't know the code): i'm not shipping logs to s3 and i'm running this on a box i built on aws running ubuntu 14.04 without using either cloudfront or the datomic transactor ami and i'm using cassandra as the datastore

architecture question: could you start by running everything on one server instance: nginx, your clojure app + datomic peer, datomic transactor, and sql storage?

then grow by breaking things out… like moving storage to dynamodb

etc etc

@chadhs: for testing, initial running it would work, but we don’t provide support for datomic configs that aren’t distributed in production. The reason being that combining processes that way impacts the stability of other processes and you need storage and the transactor to run smoothly to avoid hiccups in availability.

@bkamphaus: so at a minimum you’d want appserver, transactor, storage split

@gworley3: I’m still stuck thinking if there are any other differences I can probe. I may run an end-to-end deploy/config test with the latest version for aws metric reporting to see if anything of note comes out. Apart from that, not sure what the difference could be.

@chadhs: correct.

cool thnx, that helps

Hey, quick question about the console and licensing: I'm setting up a separate staging and production environment, and I'm considering having the console running on a dedicated server somewhere via console -p 8080 staging datomic: prod datomic:

...so I'm wondering if doing that would wind up consuming a processor license from both staging and prod even when nobody is actively using the console, or whether it only takes up a process when someone has, say, logged into it and selected "staging" from the dropdown

@timgilbert: whenever it’s running it consumes a process.

Part of the motivation is to allow developers to use the console without necessarily needing the full datomic install on their laptops

Ok, thanks @bkamphaus. So if I run it as above, it will be consuming one each from staging and production as long as the process is up, correct?

that’s correct (its connection to each as a peer)

Ok, cool. Thanks again for the info.

@bkamphaus: thanks for taking a look. i keep hoping there's something obvious that i've failed to do that would address it. fyi running version 0.9.5344

Is there a way to get the size of the the DB?

currentoor: not via the API, that I’m aware of. But you can always go look at your storage directly

@arohner: good point