Re:Riak 0.12 Released
by lylejohnson on Tuesday July 20, 2010 @03:57PM
(#1533267)
Since I didn't know:
"Riak is a Dynamo-inspired key/value store that scales predictably and easily. Riak also simplifies development by giving developers the ability to quickly prototype, test, and deploy their applications
A truly fault-tolerant system, Riak has no single point of failure. No machines are special or central in Riak, so developers and operations professionals can decide exactly how fault-tolerant they want and need their applications to be."
Re:Riak 0.12 Released
by paulsmith on Tuesday July 20, 2010 @04:35PM
(#1533358)
Is there a concise document somewhere that explains why one would choose Riak -- i.e., what unique features or characteristics -- over, say, Cassandra or another DHT-style key/value store?
Re:Riak 0.12 Released
by xal on Tuesday July 20, 2010 @08:28PM
(#1533933)
I wonder about Riak's storage efficiency. It seems like it would be ideal private S3 replacement for user upload storage. Just put a varnish in front that only allows GET requests and perhaps add/remove some headers and it should be fine exposed to the web.
Re:Riak 0.12 Released
by rb2k_ on Tuesday July 20, 2010 @04:32PM
(#1533352)
According to the IRC chan, they're also going to release the lucene based search interface to Riak.
In that case, you'd have a fault-tolerant, easily scalable datastore that allows fulltext search (among other things).
Can't wait :)
Re:Riak 0.12 Released
by roder on Wednesday July 21, 2010 @09:58AM
(#1535175)
AFAIK, Riak search is suppose to be drop-in replacement for Solr http://lucene.apache.org/solr/. So anything that wraps that API would be easy to migrate to Riak Search later.
Re:Riak 0.12 Released
by sausagefeet on Tuesday July 20, 2010 @05:31PM
(#1533512)
Anyone have any experience with Riak they could share? Riak seems like one of the less talked about DB's in the NoSQL arena, and I'm not sure if that is because it is bad or poor marketing or what.
Re:Riak 0.12 Released
by siculars on Tuesday July 20, 2010 @06:22PM
(#1533664)
it seems cassandra, mongodb and couch have gotten the lions share of the attention. of the three I would say couch is the closest to riak. Both are native erlang apps, speak http natively and use map/reduce as a primary mechanism of retrieving data.
riak is just about 1 years old from an open source perspective but is in use in a few production environments and had been in development a few yrs before going open source, afaik. What I like most about riak is the 'no special node' feature and the very close adherence to amazon's dynamo lineage. easy scale up - and easy scale down. adherence to dynamo also means you can fine tune your r/w/n values at a granular level. beyond that, the fact that you can do virtually everything there is to do in riak over http is fantastic. other access pathways are erlang client and protobuffs. another major feature of riak is that you can link keys together with link headers. admin wise, riak is very simple to get up and running either in a dev environment or production.
the drawbacks against riak are that there is no ordering of keys, no select * keys, no select range. listing all keys is expensive. you almost have to keep your own internal index of keys if you want to maintain performance over m/r jobs in any sort of realtime fashion.
im glossing over lots of finer points but if you have specific questions the irc is well manned and so is the mailing list.
Re:Riak 0.12 Released
by rb2k_ on Tuesday July 20, 2010 @07:00PM
(#1533735)
I'd be careful to compare the map/reduce in couch with the one in Riak.
A lot of the time, map/reduce in CouchDB is used to create something similar to a "view" in SQL. This is an incremental process and is automatically done for adding only changed documents to the result the next time it is called.
In Riak, map/reduce doesn't have this fancy feature.
A big plus for Riak is the fact that you can actually shard riak automatically. Couch has to be sharded manually which destroys map/reduce.
"Riak is a Dynamo-inspired key/value store that scales predictably and easily. Riak also simplifies development by giving developers the ability to quickly prototype, test, and deploy their applications
A truly fault-tolerant system, Riak has no single point of failure. No machines are special or central in Riak, so developers and operations professionals can decide exactly how fault-tolerant they want and need their applications to be."
http://blog.mozilla.com/data/2010/05/18/riak-and-cassandra-a...
http://blog.heroku.com/archives/2010/7/20/nosql/
Riak's nice for redundancy and high availability stuff, and scales dynamically by adding instances.
Anybody used it before?
riak is just about 1 years old from an open source perspective but is in use in a few production environments and had been in development a few yrs before going open source, afaik. What I like most about riak is the 'no special node' feature and the very close adherence to amazon's dynamo lineage. easy scale up - and easy scale down. adherence to dynamo also means you can fine tune your r/w/n values at a granular level. beyond that, the fact that you can do virtually everything there is to do in riak over http is fantastic. other access pathways are erlang client and protobuffs. another major feature of riak is that you can link keys together with link headers. admin wise, riak is very simple to get up and running either in a dev environment or production.
the drawbacks against riak are that there is no ordering of keys, no select * keys, no select range. listing all keys is expensive. you almost have to keep your own internal index of keys if you want to maintain performance over m/r jobs in any sort of realtime fashion.
im glossing over lots of finer points but if you have specific questions the irc is well manned and so is the mailing list.
A big plus for Riak is the fact that you can actually shard riak automatically. Couch has to be sharded manually which destroys map/reduce.