Couchbase support in Jamm!Memory

New storage

Jamm!Memory, PHP library with universal interface for best cachers (Redis, APC, Memcached), now have object for work with Couchbase - CouchbaseObject.

All features of this library are implemented in this class: tags, dog-pile and race-conditions protection, lock/acquire and others.
All code is test-covered, tested and even benchmarked. CouchbaseObject is not yet used in production, but I'm going to (see below).

Main feature of this storage is a fault-tolerance, of course.

While working on this code and testing it, I built cluster from few nodes and it was really simple. Server GUI is super user-friendly, has lot of monitoring things and control of nodes. Add new node in cluster is just  2 minutes (maybe less) plus time for rebalancing (it's going automatically and all data is available).

And next branch

Age of Jamm!Memory API is more than 2 years already. So I made branch 1.0, where currently placed all code of library, tested by unit-tests and by heavy usage in production for 2 years.

In next branch, 2.0, I will write refactored version of Jamm!Memory. API will be changed, all features will be saved, new features will be added. Some other storages maybe will be added (Riak, Cassandra).

Plans

I want to create another library, dedicated to Redis Cluster. It's in active development still, but it's time to write clients already :) It will be only key-value storage, but I hope it will as fast as existing Redis. I don't like that "Redis cluster sacrifices fault tolerance for consistence", but maybe it's just me and I don't understand something.

Also I hope Couchbase will have ability to turn off permanent writes to disk and will be able to copy data in background, so all writes and reads of data will be in memory. 
Please, support this idea by posting in forum thread.

In their forums I've created few themes and accidentally (I swear) whole column "last post" had become filled with my nickname :)

Also, I'm going to use Couchbase in production as a storage for cache, logs, sessions. This DB could be perfect for analytics (because of Map/Reduce), but it will require to write some GUI to show all these results. Currently used tool works only with RDBMS by ODBC. 

Reading Couchbase documentation

In this post I want just highlight some wondering lines from Couchbase documentation

"All Couchbase SDKs automatically read and write data to the right node in a cluster. If database topology changes, the SDK responds automatically and correctly distribute read/write requests to the right cluster nodes. Similarly, if your cluster experiences server failure, SDKs will automatically direct requests to still-functioning nodes."

"For optimal performance and speed, stores most frequently used data in RAM. Persistent storage also provided to survive system downtime and for re-population of RAM."

"For example, a player making 10 game state mutations in 5 seconds, such as planting 10 flowers in 5 seconds, will be compressed by Couchbase automatically into just one queued disk mutation."

"To extend our application with new user attributes, we simply start storing additional fields at the document level. Unlike traditional relational databases, there is no need for us to have server downtime, or database migration to a new schema."

"You may wonder how effective it is to run query your view if Couchbase Server will run it on every persisted document in the database. But Couchbase Server is designed to avoid duplicate work. It will run the function on all documents once, when you first query the view. For subsequent queries on the view Couchbase Server will recompute the keys and values only for documents that have changed."

Impression

Finally I've read that Couchbase Developer's Guide 2.0. 

Very good example of documentation. Not the best, but very good. Many DB sites can learn from Couchbase how to write good documentation.

In some articles too many words and trivial examples (even with pictures), but sometimes amount of examples in live code is just not enough. For examples, after reading whole manual, I still don't know how to create "view" and how send it to server. Some file saving... But I know how to write functions in view and even best practices about it :)

Of course, I'm very impressed about Couchbase at all.

Ability to write map/reduce functions in JavaScript is amazing.

Auto-updating results of view is awesome.

Finally I know how my idea about select via callback function should be named - "map function" - I knew I'm reinventing the wheel :)

Tomorrow

I'm planing to add support of Couchbase to Jamm!Memory, I'm planing create there branch 2.0 (for "cas" and some other things in all cachers and to rename some methods to more familiar get/set).

I'm going to learn Ruby and try to write project in Ruby. Examples in Couchebase documentation shows that Ruby can be readable even easily than PHP.

I really think that Couchbase is the best in the NoSQL world at this moment. But maybe I just don't know all of them :)

Looking at Cassandra DB

Few days ago I was asked "what do you think about Cassandra?" At that moment I'd read only few short notes about this database, all these notes was pretty old (around year ago) so there was not many useful or attractive information. But I've decide to check what has been changed with this DB. And... it was really interesting.

Already after week I've read book Cassandra: The Definitive Guide and lot of information from StackOverflow, SlideShare and googled articles. Book is very outdated, most of useful information you can find in Datastax's Cassandra documentation. To update your knowledge after reading that book, I recommend to check list of articles at http://www.datastax.com/dev/blog (for example, SuperColumns were replaced by Composite Columns and Thrift was replaced by CQL3 and binary protocol).

And what is so interesting in Cassandra?

First of all - fault tolerance

I'm sure it's very important for clients loyalty to see working service. They shouldn't see "server is down due maintenance" message never - it's unacceptable and it's shame for every service.

Most of web-projects currently use MySQL as database and it's not so easy to achieve real fault tolerance. Master-slave replication is not enough - what will you do when master-node will be rebooted for software/hardware upgrade or will be not available in case of network issues? You will panic and try to switch some slave into master mode or restore master. It's a downtime, in good case - few minutes, in bad - few hours. Users loyalty, reputation of service, money, time, sleepless night...

It's good that we have Galera for MySQL. Update: Don't use Galera, it's painful.
But Amazon Dynamo, Cassandra, Riak were build with idea of decentralization as the main idea so it's not "workaround" for database, it's intended behavior. That's why I personally trust Cassandra's fault tolerance more than MySQL's with Galera.
Just read about all possibilities and settings of replication in Cassandra and you will see that there is a lot of features for great fault tolerance. Cross-datacenter ring native support is awesome.

Slide from eBay's presentation, they have articles about Cassandra best practices also.

Second - performance

I spent whole day to find this page with benchmarks graphics of Cassandra, MySQL, Riak, HBase, MongoDB. But here I found very important information - performance of Cassandra is very good, not just "good enough" but exactly very good.

Ready to use

We, modest programmers of high-level web-languages, are always in a dependence of "clients" for compiled software.

It was not surprisingly to find an existing PHP-client for Cassandra, PHPCassa, but I don't like static methods, hardcoding and global variables, so I don't want use code with so ugly lines as "$GLOBALS['\cassandra\E_IndexOperator']" (it's line from PHPCassa code). Other clients have lot of static method also.

I thought there is 2 variants - write own client (as wrapper for Thrift API) or find existing client with most clear code. But... today I found very cool thing - Cassandra PDO driver for PHP! And there is SQL-like syntax (CQL) It's amazing and now I'm deploying few Linode VPS to try Cassandra PDO.

I'm inspired and that's what I wish to you :)