CloudTran Product Features

Great Features Every Developer Needs

CloudTran provides the features and capability every developer needs to run scalable web applications. Fully integrated with data grids including Oracle Coherence and GigaSpaces XAP, CloudTran gives you out-of-the-box persistence to popular data stores all within the same transactional context as the data grid. And, with sharding built into the CloudTran product, you'll be able to scale-out your data store level as well.

CloudTran Features

Reliable back-up for your favorite in-memory data grid using industry-standard data stores.

Data grids represent essential technology for today's web-scale applications. By providing unified access to shared memory across loosely-coupled computers, data grids offer the ability to store entire databases in memory resulting in the best performance possible. CloudTran enhances the capabilities of data grids with asynchronous, transactional persistence to sharded data stores.

  • Creates and manages a write-behind queue for all inserts, updates, and deletes made into the data grid.
  • Asynchronous processing to the data stores to keep applications running at in-memory speeds.
  • Support for Oracle Coherence and GigaSpaces XAP data grids and Oracle, SQL Server, MySQL databases; also supports JDBC.
  • Support coming soon for MongoDB NoSQL data store.
  • Handles fail-over recovery for grid nodes that may go down.

Fast distributed transactions with full ACID properties and without 2-phase commits.

CloudTran adds ACID symantics to all your transactions, making it possible to write complex transactions that span in-memory data nodes and sharded data stores. Importantly, CloudTran does not use the traditional 2-phase commit protocol to coordinate transactions, which is often too slow for operational applications. The net result is you get turbocharged applications that run at the speed of the in-memory database, but with all the transactionality and persistence of traditional data stores.

  • Rock-solid, ACID transactions across grid nodes, providing greater flexibility for data placement and less worry about co-locating transactional data all in the same node.
  • ACID transactions when persisting data from the grid to back-end data stores; eventual consistency ensures all data will be reliably backed-up even if one or more database shards is temporarily unavailable.
  • ACID transactions across data shards, across heterogeneous storeage systems, and over LANs/WANs, public clouds, private data centers, and hybrid architectures.
  • Read commited isolation level.
  • Transactions commit first in the data grid for the highest performance.
  • Uses proprietary logging and commit protocol to avoid 2-phase commit.

One view of data no matter how many pieces it's in or where it's stored.

CloudTran allows you to shard back-end data stores for scale-out performance. Instead of funneling all transactional data to a single database, you can spread data over multiple database to improve performance and support extremely large data sets. Multiple, heterogeneous data stores can all be included in a single logical view giving you the flexibility to chose the right product for the job. For example, in an ecommerce application, structured data such as customer orders might be stored in several relational databases, but catalog data could be stored in a NoSQL system. CloudTran will transparently and transactionally interact with these disparate systems.

  • Unlimited number of shards supported.
  • Shards can be located on the same node, over LANs, and over WANs.
  • Shards can be in public clouds and/or private data centers.
  • ACID transactions across shards without using 2-phase commit protocol.
  • New shards can be added while system is in use to add dynamic scalability.
  • Support for heterogeneous sharding (i.e., SQL & NoSQL in the same transaction).

Build a data layer that's here today, and... here tomorrow.

When it comes to the data layer, CloudTran provides an out-of-the-box architecture that removes database bottlenecks and provides virtually unlimited scalability. CloudTran will coordinate transactions that span as many nodes as you need to achieve the performance you want. Have a growing user load? Make a quick configuration change, and without any modifications to your code, CloudTran will automatically take advantage of additional computing resources for both in-memory data as well as persistent storage.

Whether you need high-end performance today or not, CloudTran is a great place to start for every application. Our developer tools help build best-practice data architectures that won't let you down as needs grow and change. Working within a distributed system is very different from the days of client/server architecture, and CloudTran is the first toolset for seamless development of this new style of application.

© Copyright 2012 CloudTran, Inc.| All Rights Reserved.