Creating a Data Management Strategy

Creating a Data Management Strategy

Data management will be the most challenging part of your hybrid cloud project, but it’s also the key to avoiding the creation of data silos as you move your applications to hybrid cloud.

It’s often tempting for enterprises to forgo establishing a data management strategy, but they shouldn’t. Ignoring this critical element increases the risk of creating another data silo with redundant data from other applications that need to be kept synchronized by multiple players.

For example, suppose you have several databases that contain customer information and each database allows applications to create, read, update, and delete (CRUD) customer data? To support this, there is likely a feed or sync program that keeps the customer databases in sync. In this situation, if you decide to build microservices in the cloud, it’s important that you resist the temptation to create another database—and therefore another data silo—that contains customer data on which applications perform CRUD operations.

Instead, you should design microservices that perform CRUD operations. Each micro service operates on, for example, customer data in one database. This microservice becomes the conduit for updates for every other application that updates the same data in another database. There are many ways for doing this, but that is the subject of another white paper. The transition for an application to use microservice to obtain data updates will be different for each application. This transition has a higher impact on systems that are costly to change, such as a COBOL application on a mainframe.

If you must create a new database in the cloud, first consider the corporate data view. Identify company systems that already have the data that you plan to store in the cloud and use these systems to source the data. If your company has a conceptual data model, use it in the design of the new database.

DataStax Enterprise (DSE) has a unique masterless architecture, which is natively architected for hybrid cloud, allows you to build a consistent data layer across on-premises, hybrid cloud, and multi-cloud infrastructures. It eliminates data silos and accelerates hybrid and multi-cloud application deployments.

As a hybrid cloud database, DSE provides all the operational, or transactional, data in a single logical layer for global data availability. The data is highly distributed, always available, and highly secure, delivering the data and ingesting massive amounts of writes in real time. With this architecture, enterprises can deploy part of the data layer on premises, part of it in one public cloud, and part of it in another public cloud. Enterprises can also move the data in and out of any cloud at any time with no downtime or rewriting of their applications.

Create A Security Strategy

An IT organization always needs to comply with its enterprise’s security policy, and this extends to hybrid cloud. For example, the enterprise may be using AWS in conjunction with Azure or GCP and wish to keep data associated with its retail store business unit in Azure and GCP but not in AWS.

This Is Easily Accomplished With DSE

DSE includes advanced security, which fortifies the database against potential harm from deliberate attacks or user error. DSE has advanced mechanisms for authentication and authorization; encryption of data in-flight and at-rest; and data auditing.

DSE is also compatible with various partner security solutions to meet industry-specific and other advanced requirements. It leverages enterprise standards to integrate cohesively with existing technology such as Active Directory (AD), Lightweight Directory Access Protocol (LDAP), Kerberos, Public Key Infrastructure (PKI), and Key Management Interoperability Protocol (KMIP).

Monitor Your Distributed Database Via a Single View

A best practice for the building and deployment of any application—whether simple or complex or deployed in a private, public, or hybrid cloud—is to create, test, and maintain (by keeping it up to date after going live) an operational runbook.

This extremely important document contains critical operational procedures, including monitoring, security, access control, and configuration. The operational runbook is used not just for the database but also for microservices, backup and restore, event handling of all types (such as if a node goes down or a cloud region is not available), SLAs not being met, background jobs not running, how to handle failures (in detail), and upgrades.

However, monitoring database workloads, activity, auditing, and performance metrics throughout an enterprise is very tricky and becomes even more complicated in a hybrid cloud environment, particularly if it includes public clouds.

DSE configured in a hybrid cloud allows for monitoring and distributing the database across clouds in a single view. The DSE Metrics Collector is built on collectd, a popular, well-supported, open source metric collection agent. With over 90 plugins, you can tailor the solution to collect metrics most important to your organization. When DSE starts, it automatically begins sending metrics and other structured events to the DSE Metrics Collector, where the frequency and type of metrics collected are easily configured. These metrics can aid in the identification of read and write performance latency that is outside of a defined SLA. The aggregated metrics can also be sent to many metrics visualization and monitoring tools such as Prometheus, Graphite, Splunk, and Grafana.

Of course, you can use DataStax OpsCenter as your visual management and monitoring solution for DSE. OpsCenter provides architects, database administrators, and operations staff with the capabilities to intelligently and proactively ensure their database clusters are running optimally across multiple clouds with a single view. Furthermore, DataStax Lifecycle Manager allows enterprises to easily configure and upgrade DSE in hybrid cloud environments, with everything exposed via RESTful API so that it’s fully compatible with all tooling.

Conclusion

Most of today’s enterprises are planning to deploy hybrid and multi-cloud data architectures, if they’re not doing so already. But this transition often comes with drawbacks, such as data silos and data governance issues.

DSE makes it easy to fully exploit hybrid and multi-cloud environments without having to rearchitect cloud applications.
A masterless architecture can provide consistent data fabric between on-premises and cloud-based resources, without
compromising on the availability, scalability, and security required for modern applications.

The distributed NoSQL database allows you to implement a successful hybrid or multi-cloud strategy because it:

  • Offers distributed database management across
  • multiple clouds
  • Eliminates single points of failure in your hybrid or
  • multi-cloud deployment
  • Helps you optimize your data management strategy
  • Enforces your security strategy throughout your hybrid cloud
  • Lets you monitor your distributed database with a single view of your hybrid cloud

With these five key abilities, any enterprise can take full advantage of everything hybrid clouds have to offer. 

Tagged

Hybrid Cloud Multicloud OpsCenter

Open-Source, Scale-Out, Cloud-Native NoSQL Database

DataStax is scale-out NoSQL built on Apache Cassandra.™ Handle any workload with zero downtime and zero lock-in at global scale.

Get started for free
Astra Product Screen