NoSQL database company Couchbase is adding new scaling technology in its next major release of Couchbase Server to increase application performance as well as reduce costs. The new Multi-Dimensional Scaling capability provides the option to isolate database query, index and data services, so that hardware resources can be independently assigned and optimized on a per node basis, as application requirements change. This improves performance, reduces hardware costs, and enables enterprises to support a much broader set of applications with a single database. Couchbase Server 4.0 with Multi-Dimensional Scaling is planned for release in the summer of 2015.
Learn how real users are building globally available applications using NoSQL. Register for a special DBTA webcast on Thursday, March 26, 2015 at 11 am PT / 2 pm ET - "Enabling High Availability for Your Applications with NoSQL"
Describing the ongoing evolution of enterprise NoSQL adoption, Bob Wiederhold, CEO, Couchbase, said it is occurring in three phases. Phase 1 was about grass roots developer adoption with developers downloading the code, installing it in order to explore the technology and start using it on lightweight applications. Phase 2 kicked off in early 2013, as enterprises began using NoSQL in increasing numbers under mission-critical applications that they are operating at scale, which in turn drove a put a higher priority on scalability and high performance to support those applications. Phase 3, Wiederhold said, will be starting next year when enterprises begin deploying NoSQL more broadly. By then, they will have developed confidence in NoSQL under mission-critical business applications.
“We think Multi-Dimensional Scaling is very important for success in phase 2 as well as in phase 3. This is a big announcement of a fundamental capability,” said Wiederhold.
Until now, the query, index, and data services of a database have typically shared the same hardware and competed for resources. This can result in expensive over-provisioning of resources as well as degrade application performance due to resource contention. Couchbase Server 4.0 with Multi-Dimensional Scaling will seek to eliminate these problems by allowing organizations to run services on separate hardware and assign right-sized servers for each service. Organizations will have full flexibility to use Multi-Dimensional Scaling or to deploy standard distributed scaling with all services running on the same hardware, and can change on the fly.
Benefits of Multi-Dimensional Scaling
Multi-Dimensional Scaling will benefit database deployments in tasks such as:
- Querying, whichis a CPU-heavy operation. With Multi-Dimensional Scaling, enterprises will be able to isolate the query service and assign to it a small set of low cost commodity boxes or a large box with more computing power. This will result in faster queries and avoid impacting other services.
- Indexing, which is disk-intensive. Multi-Dimensional Scaling will allow enterprises to isolate the index service so that index operations are performed only on the assigned hardware. This will improve index performance and enables enterprises to maintain any number of indexes without degrading read/write or query performance.
- Basic data access, where fast reads and writes are the primary requirement. With Multi-Dimensional Scaling, enterprises will be able to isolate the data service on low-cost boxes to maintain sub-millisecond read / write operations with no degradation from query or index services.
“Today, all of these tasks typically share the same underlying hardware - and you run all of those on the same servers and so, as a result, they compete with one another for memory resources, CPU resources, and disk access resources,” said Wiederhold. The result, he said, is that either organizations have to over-provision all of their servers in order to be able to handle all the of the reads and writes, indexing, and querying, which can be costly, and/or they end of degrading their application performance because there is resource contention. For example, if they are doing intensive index-building or are running intensive queries, those processes could slow down their basic reads and writes.
“That is a trade-off that customers really don’t want to have to make, but right now they are forced to do so because everything is running on the same servers,” said Wiederhold. “Multi-Dimensional Scaling eliminates these needs because now you are going to be able to run your indexes and queries - if you want - on a different set of servers. You can scale to meet your requirements for indexing and querying base d on the number of indexes you are building, and on the number and intensity of the queries you are running, and you can run them on a separate set of servers from the basic data access to just serve your basic reads and writes.”
Go to www.couchbase.com for more information.