Hacker News new | ask | show | jobs
ArangoDB 3.2 GA RocksDB, Pregel, Fault-Tolerant Foxx and Satellite Collections (arangodb.com)
124 points by bjerun 3258 days ago
12 comments

Very good news! :) RocksDB engine is a very big step of ArangoDB growth! I use ArangoDB in some projects and it's absolutely fantastic: very powerful AQL queries, very-very fast, good optimizations.

I checked alpha & beta versions of 3.2 and the improvements are really amazing :)

The graph layers look wonderful, I'm happy to see they have progressed so far!

The combination of Arango's new capabilities with Graphistry should be interesting. So, if Arango graph users are interested in trying GPU visual graph analytics for looking at more data at a time, our team is happy to share access to http://github.com/graphistry/pygraphistry . Likewise, investigation teams digging daily into event & entity data (security, fraud, patient journey, ...), we are piloting our visual playbook automation & interactive investigation layer with medium & large enterprises, and we'd love to chat about that as well. Either way, ping info@graphistry -- this sounds like a great match.

Jan from ArangoDB here

Sounds cool, let's talk :) feel free to ping me jan.stuecke at ...

Congratulations to the ArangoDB team ! With ArangoDB you can create powerful applications using a multi models database and an API builder called Foxx. I'm building all my apps now within ArangoDB/Foxx no more needs of server apps.
Thank you!

One thing the article actually forgets to mention is that 3.2.0 also is the official release for the new HTTP API for managing Foxx services: https://docs.arangodb.com/3.2/HTTP/Foxx/Management.html

We're hoping to backport most of it to 3.1 and already support it in the latest release of the JavaScript driver. This should make automated deployments considerably easier. You can literally install, configure and distribute a service with a single line of `curl` now.

Full disclosure: I'm one of the core developers working on ArangoDB Foxx.

Semi-related to Foxx, is it now possible to add/remove coordinators dynamically to cluster? In other words, can we now scale out Foxx services easily if load increases?
It's trivial to add additional coordinators and removing coordinators no longer has a chance of affecting Foxx because the authority for the deployed services now lies with the database servers rather than the coordinators -- the coordinators are now fully disposable.

Be sure to check out this short demo of the Cluster Starter:

https://www.arangodb.com/fault-tolerant-foxx/

Never mind. Just seen your reply and link on other comment. Looks very promising, cheers.
I find fox quite convenient too.
Great news, I've been using ArangoDB for over a year and this upgrade is great, ArangoDB is shaping up to be a 'Giant Killer'.

The new automated deployment options for Foxx as well as the better memory management with RocksDB are highlights for me.

Thanks! I hope we will get some good feedback for all our new features
Congrats, ArangoDB team!

Reading through the release notes and related blog posts lead to several wow's and nods.

Especially the satellite collection feature that allows local joins in cluster mode sounds promising. Can't wait to learn from successful real world implementations.

Hope to find some time to get used to all those new features announced today.

One point that is still missing here and I have been very excited about is trying out Pregel! Distributed graph processing is something I have really always wanted. Thanks a lot to the team for adding this.
Awesome news! Go Arango team
I wonder if they improved Cluster setup with this release. If you tried without DC/OS, it was painful experience. Arangodb Starter was definitely a step in right direction.
Hi, it's Frank from ArangoDB. We have now included the starter in the package. It is now possible to start a cluster with a single command line. For example, to start a test cluster on a single machine, "arangodb --starter.local" is all you need to type. Starting on 3 machines requires a "arangodb" on the first machine and "arangodb --dataDir=./dbX --join serverX1" on the others.
Thanks Frank. I wonder if Web Fronted is in order akin to Counchbase setup process. Ultimately CLI tools are of course the king of the Ops world when maintaining clusters. Simple Web app can go long way though for attracting new users. First impression matters. My 2c.
ArangoDB starter is now bundled with the distribution and can start single server instances as well as cluster. It can use Docker or normal processes.
Have you had a look at the fault-tolerant Foxx demo?

https://www.arangodb.com/fault-tolerant-foxx/

Although it showcases fault-tolerance in Foxx specifically, it's really a demo of setting up a cluster with Arango Starter and we wouldn't have been to make Foxx so resilient in 3.2 without the hard work of the cluster team providing the stability underneath.

Disclosure: I'm a core contributor for ArangoDB Foxx.

The new release looks packed full of features - especially excited about Keylines and encryption.

FYI, a minor copy edit suggestion: "HIPPA", should be "HIPAA".

-https://en.wikipedia.org/wiki/Health_Insurance_Portability_a...

Thanks! Yeah, some features we wanted in 3.2 ended up taking longer to polish than anticipated so the rest of us were able to work on features that would have otherwise been scheduled for 3.3 I guess :)

I've forwarded your find to the marketing team.

Congrats on shipping! I wonder if RocksDB could be the right choice for small to midscale IoT scenarios? Anyone has an opinion on that?
I think yes. The RocksDB engine is much better than the old engine in situations in which the stored data is larger than the available RAM. It all depends on the concrete queries issued but in general I would say: "Definitely yes!"
I use ArangoDB for all my projects, big, medium, and small. Happy to see great new stuff.
Would like to hear who used it in production and what type of scale that DB is handling.
Luca from ArangoDB here... feel free to reach out to me at first at arangodb.com... we've lot of production users and customers
Why not just talk here where there are a lot of people share the same interests than redirect them to your website?
Good point... the public references are here: https://www.arangodb.com/why-arangodb/case-studies/ just wanted to offer a more personal approach... also, some of the people shared their reviews on https://www.g2crowd.com/products/arangodb/reviews
Amazing, this release has many features that I was waiting for, I already switch my project from MongoDB to ArangoDB, but I still missing realtime features, or a least a tailable cursor like in MongoDB... But even so, it's worth it!
For single node setups you could try arangochair: https://www.arangodb.com/2017/03/arangochair-tool-listening-...