This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2016-12-02
Channels
- # adventofcode (5)
- # arachne (2)
- # bangalore-clj (1)
- # beginners (8)
- # boot (195)
- # cider (28)
- # cljs-dev (35)
- # cljsrn (4)
- # clojure (295)
- # clojure-brasil (5)
- # clojure-gamedev (2)
- # clojure-greece (2)
- # clojure-korea (13)
- # clojure-russia (60)
- # clojure-spec (58)
- # clojure-uk (92)
- # clojurescript (31)
- # clojurex (4)
- # css (1)
- # cursive (13)
- # datomic (40)
- # devcards (2)
- # emacs (17)
- # events (1)
- # flambo (3)
- # garden (9)
- # hoplon (31)
- # jobs (3)
- # klipse (1)
- # lein-figwheel (1)
- # london-clojurians (1)
- # luminus (2)
- # mount (36)
- # off-topic (13)
- # onyx (8)
- # pamela (1)
- # pedestal (1)
- # planck (3)
- # proto-repl (16)
- # protorepl (11)
- # re-frame (78)
- # reagent (4)
- # rethinkdb (6)
- # ring-swagger (1)
- # specter (8)
- # untangled (10)
- # vim (1)
is a separate partition for audit logs a good idea? I can easily store that on S3 as well, but why wyould i want to do that? if in db i can easily retrieve, query etc interesting audit infromation. Separate partition would also help in log rotation? And as one can see i do not work on high volume system by todays standards 🙂
what would go in your audit logs that you couldn’t just put on your transaction entities directly, @karol.adamiec ?
hmm, say user A tried to log in and failed on auth checks.
ok, so logging events for user gestures
then yes, a separate partition is a good idea
we do this, but not the partition. planning to re-build our db and put them into its own partition
ok. if i ever need to rotate the logs (keep a year only) does partition give me any advantage?
in that case, you’ll want a separate database, and you’d make a new database every time you rotate
you don’t want to put lots of data in that you plan to take out later. datomic isn’t really designed for that
yeah, most likely i will never need to rotate that anyway. It is not a huge system….
separate partition seems like a good compromise for now…
at least will keep audit/log noise out of user space.
indexing is per partition as far as i understood
@timgilbert Thanks for the catch! I will update
Is it a bad idea to create identities while adding data? I have a case where I query external API that returns things like {:id “foo” :badges_collected [”badge_bar” “badge_baz"]}
I know that the set of badges doesn’t change very often but there’s a lot of them and I don’t have a decent way of extracting them all. I want to express the badges as enum so I’d transact something like: [{:db/id … :entity/id “foo” :entity/badges [:badge/bar :badge/baz]}]
And I would need to transact [{:db/id … :db/ident :badge/bar} {:db/id … :db/ident :badge/baz}]
first for every entity I’m about to save.
@nooga I’m going from memory here, but I believe idents are cached on each peer, so the more you have, the more memory you use. That might limit you.
yep, and i think there’s a 32k limit on the overall count, @nooga and @bhagany
(i think)
yeah, I think your transact-as-you-encounter strategy is fine for that low of a number
Hi all, really want to get started with Datomic. Any good tutorials or books out that that I’m missing? What’s the best way to start learning this?
This is probably a n00b question… but I’m trying to use the new client and I’m having some issues. If I use the REPL packaged with datomic, the connection works great. However, if I use the REPL in my own project… It gives me the
#:cognitect.anomalies{:category :cognitect.anomalies/unavailable, :message "java.net.ConnectException: Connection refused”}
Anyone have a hint to what might be wrong?@aramz it wouldn't hurt to get a mentor for a few months to make sure you're on the right path. what things do you know about Datomic right now?
I have gone through the training videos on http://datomic.com