Hot Databases

databases
Tags: #<Tag:0x00007f493ffb6cf8>

#21

Check the new features for PostgreSQL 9.5: http://www.postgresql.org/about/news/1636/

PostgreSQL is great also in terms of cluster management for scalability and replication. Futhermore it has great algorithms for server- and session-level caching for performance.

Check my post about a PG example: https://www.linkedin.com/pulse/applying-some-good-open-source-savvy-hacking-big-data-moriondo


#22

I’m looking into Citrus Postegre Extension, for distributed database.

It is apparently quite simple to deploy…


#23

Also give a look at ArangoDB - i will write a Adapter (Ecto) soon :slight_smile:


#24

Power tools: Sorting through the crowded specialized database toolbox by arstechnica


#25

http://softwareengineeringdaily.com/2016/04/13/cockroachdb-ben-darnell/


#26

I really liked this article. So refreshing to get some form of reasoning for the database choice rather than the common Versus articles.


#27

Does anyone have any experience with graph databases? Namely, Neo4j? Could you share your story or opinion? :slight_smile:


#28

Shame that RethinkDB is shutting down…

https://www.rethinkdb.com/blog/rethinkdb-shutdown/


#29

Well it’s a shame but at the same time it’s a strong illustration of why we need to give careful consideration to how sustainable a project is before using it in production apps. If 90% of development is coming from salaried dev’s eating through a pile of VC money and project has no reasonable monetization strategy …


#30

http://www.datomic.com/ , a “functional” db. Rich Hickey has some items on youtube about databases also, f.e. https://www.youtube.com/watch?v=Cym4TZwTCNU


#31

Yes datomic is nice, but sadly is close source and mostly used by Clojure.
Also pricing is not perfect.


#32

I am interested in eventstore and also couchdb.

I found the former while looking at datomic and the ideas of event sourcing. Couchdb is interesting having recently just release 2.0

However I think that the database free architecture is a really interesting idea and I would be interested in hearing anymore real world stories of deploying something like that.


#33

I was using CouchDb in my javascript POC implementation of ibgib, because it integrated very nicely with PouchDb on the client side. Both were very nice document storage dbs. CouchDb, being written in Erlang, was actually one of the things that led me to Erlang and ultimately here to Elixir!

I had also previously used eventstore, and it was a very powerful, highly optimized event sourcing back end. I was not savvy enough to get it set up consistently on a Windows server at the time though. Perhaps it is easier now, plus I’m (slightly) more used to linux.

Interestingly, this thread has helped me realize that ibgib’s current approach has properties very similar to a mix of event sourcing, neo4j, IPFS, and “functional” code-as-data. But I digress.

Both CouchDb and EventStore were great tools in my (non-production) experience.


#34

The talks by Greg Young on CQRS are good that’s how I found EventStore but the thought of having infra. type stuff running on windows or depending on Mono I find very unappealing.


#35

I am wondering If can achieve the same result as evenstore database using Kafka if so I will go for Kafka
http://www.confluent.io/blog/event-sourcing-cqrs-stream-processing-apache-kafka-whats-connection/


#36

Kafka is not a permanent store


#37

#38

Not heard of that one before @mkunikow - not sure the name is a good choice tho, apart from being a little odd/memorable :lol:


#39

Hmmm, I gotta say it was very eye-catching to me when I had previously seen it, thinking of data surviving like cockroaches. :thinking:

Regardless any of the database’s individual merits though, that is a fascinating blog post about stability problems with scaling. :thumbsup:


#40

From my pov right now, there is only a couple one i would use at work

  • Postgres is my default for anything Relational. Stable, move fast without breaking thing, quite well.
  • I have to deal with a tons of Time Series (both for metrics and for market data) so Influx did quite a splash. It is only nearly the only one in that are sadly
  • For KV, Riak is still rock solid, have great CRDT and works well, with good doc. But you need to read it to have proper conf.
  • no database? :stuck_out_tongue: i use quite a lot of ETS, CSV, etc etc
  • Rethink is on my list of distribute document DB, but to be honest, i still need to find a good one…
  • Datomic is still around and saying “use me use me” but i still have to have a need for it…