Most organizations make decisions based on data. In recent years, the amount of data that these organizations can access to help make those decisions has exploded. More data, though, won’t help unless the organization can analyze that data. It became obvious that legacy applications could not scale to meet the challenge. Modern applications, like Cassandra, MongoDB and Spark, emerged to fill the gap. The problem is this new generation of applications lacks the traditional data protection best practices.
Do Modern Apps Need Protection?
A common misconception is that modern NoSQL and Hadoop environments don’t need backup. Most of these applications have some sort of replication function built into them, and they automatically create three or more copies of data as it changes. This protection design though, only protects against media failure. Since it happens in real-time, data corruption or user error will result in a near-instant copy of the problem in all replicas with virtually no ability to rollback. The risk of a failure, whether caused by human error or application corruption, results in data loss potentially costing the organization thousands, if not millions, of dollars.
What IT Needs
IT needs a way to protect this data from both internal mistakes and external cyber-attacks. The problem is, most legacy data protection solutions have no direct support for these platforms, forcing IT to cobble together workarounds. Additionally, they need capabilities to improve their ability to migrate, test and manage data sets.
These applications need more than just backup and disaster recovery. They also need a complete suite of services including migration, test/dev management, archiving, cyber-attack detection and recovery time prediction.
Introducing Imanis Data
Imanis Data is a data management software company focused on modern application environments like NoSQL and Hadoop. The Imanis Smart Storage Cluster is a scale-out software solution that runs on commodity hardware. It leverages Hadoop as its underlying architecture, which gives it excellent scale and parallel processing support. The Imanis Data solution provides support for a wide variety of applications, including Cassandra, Couchbase, MongoDB, Azure HDInsights, Cloudera, Hortonworks, Spark, Vertica and many others. All of these applications can backup to a single Smart Storage Cluster. No agents are required on the application servers themselves, making implementation simple with minimal overall impact. After the initial backup, updates to the backup copy are incremental and rapidly completed.
Imanis Data provides multi-stage optimization, a challenge in the typical modern application architecture since most of these environments optimize storage through compression and other techniques. Imanis Data implements a post-process optimization technique, which stores data in its original format for a period, then optimizes the protected copy. First, the data is uncompressed, followed by elimination of redundancies (deduplication), then re-compressed and erasure coded to protect against media failure while minimizing disk capacity requirements.
Imanis Data also supports a variety of targets on which to store protected copies including; on-premises DAS, SAN, NAS and object storage, as well as cloud storage from major providers like Amazon and Azure. The solution also supports the tiers within cloud storage, so it can send protected copies to Amazon S3 for example and then age copies to Amazon Glacier, based on policy.
Imanis Data is more than just a data protection solution; it can also assist with cloud migration thanks to its multi-cloud capabilities. Test/dev can leverage the Imanis solution since it has the ability to instantiate snapshots of protected data and its cluster has the compute resources to drive the test development effort. The solution can also archive older data sets to limit on-premises data growth. Finally, it has a machine learning engine that enables a variety of specialized applications. For example, “ThreatSense” determines if data is under attack from ransomware. It accomplishes this by detecting, among other things, the rate of change within the protected data set and comparing it to the historical change rate.
In the future, the solution will further leverage its Hadoop foundation to run machine-learning analytics to determine actual RPO and RTO of protected applications.
The next generation of applications is now the current generation. Data centers of all sizes are implementing applications like Cassandra, MongoDB, Spark, Couchbase and others to generate customer insights, detect fraud, improve products and operations as well as personalize the customer experience.
These applications are in production and need protection just like any other production environment. But limited vendor solutions and DIY fall short in providing sufficient levels of protection. Lack of protection exposes the organization to millions of dollars of potential revenue loss if the application goes down. Implementing solutions like Imanis Data’s Smart Storage Cluster is now an imperative.