Hazelcast Enables Open Government to Meet Peak Web Demand

Industry

Government

Product

Hazelcast Platform

Making government data available to the public online is a critical and ever-growing requirement of becoming an open government. Everything from title records to business names and related registration data to census information, as well as information about laws, regulations, and general services offered by the government, are examples of this critical government service.

The responsibility for this service is either fulfilled by the government departments themselves, or they have enlisted the help of a third party to manage and publish this information for them. There are strict Service Level Agreements (SLAs) around the provision and availability of this data. Typically this data is made available to the public through web pages, web portals, or APIs. Due to the large volumes of data and the strict SLAs, without the proper software solution this effort can be costly, time-consuming, and inefficient.

Hazelcast In-Memory Data Grid (IMDG) provides a unique set of characteristics that make it the perfect choice when meeting the needs for publishing and maintaining the availability of government data to the general public. Hazelcast is regularly seeing entities turn to Hazelcast IMDG when their existing systems are unable to deliver on their requirements. It’s not just older technologies such as Oracle databases that are struggling to cope; newer approaches using NoSQL databases are also failing to live up to their early promise.

Why Most Datastores Fall Short…Why Hazelcast IMDG is Ideal…

Relational databases are the typical systems of record for government records. However, they cannot scale to the massive demands of open government.

Hazelcast IMDG can cache the relational database, sitting in front of it, serving data held in memory at sub-millisecond speeds, while at the same time relieving the workload from the relational databases. The results are faster retrieval of data and SLAs that are achieved, even during peak times like tax season. Hazelcast IMDG gives huge headroom to absorb large spikes around lodgement dates. If you already have your data in a relational database, you can add Hazelcast as a database cache without needing to perform any rearchitecting, preserving your investment.

NoSQL databases can scale out but need help to meet tight SLAs.

When used in front of a NoSQL database with “read-through” and “write-through” enabled, Hazelcast IMDG allows for scale-out of the persistence tier, and delivers very high performance to meet SLAs. Hazelcast in front of Apache Cassandra, Couchbase, or MongoDB gives you a scale out persistence layer combined with sub-millisecond speeds. The combined solution can scale out to handle the largest workloads.

In-Process caches such as Ehcache and Guava have high performance but do not scale beyond a single node. As the system scales, the benefits reduce proportionally.

Hazelcast IMDG is a distributed cache. Once a piece of data is known to Hazelcast IMDG it can be served without needing to consult the system of record again. The result is a scalable system designed to handle millions of concurrent users and the largest workloads, all while maintaining high performance.

Customer Success Story

Office of the Revenue Commissioners (Irish Tax Office)

The Office of the Revenue Commissioners have a legal duty to make taxation data public to help taxpayers make judgements in relation to tax issues. They have SLAs for website response and also for total time taken to find information requested. The entire software stack is aimed at meeting these requirements. The data is held in HDFS (Hadoop File system) and is stored in Hazelcast IMDG for fast retrieval. The entire software stack is open source including Hazelcast.

Using Hazelcast IMDG the Irish Tax Office have met their SLAs. They have scaled their system using more than 300 nodes of Hazelcast IMDG as total in-memory platform across all of their applications.