Why ACI is the best platform to run Datos IO + Cassandra Hyperscale Application – Multi TB data backup in the order of hours

Business growth, digital transformation, and business innovation are propelling enterprises to adopt next-generation, high-value, data-centric applications such as real-time analytics, operational intelligence, and machine learning. To support the data requirements of these high-volume, high-ingestion-rate, next-generation applications, enterprises are rapidly adopting massively scalable and non-relational databases such as Apache Cassandra, MongoDB etc.

To deploy web scale systems leveraging NoSQL databases like Cassandra, an easily-scalable infrastructure and a reliable backup strategy is paramount. Cisco ACI enables zero-touch addition and removal of Cassandra nodes with an intent- and policy-based language, while DatosIO RecoverX provides reliable on-premises and cloud-based backup and recovery. In this blog, I will take you on a quick tour of how the three technologies together provide a reliable, scalable, and recoverable data center solution for modern applications.

Cisco ACI, Cassandra DB and Datos RecoverX: Better Together

With the combined power of Cisco ACI, Datos RecoverX, and Apache Cassandra, applications can expect superior performance, deep application insight, and a modern backup strategy. Cisco ACI’s application-profiles model the pieces of an application into endpoint groups (EPGs), or tiers. The profiles use contracts to specify which pieces can talk to each other, and they use filters to specify what they can talk about (ports, protocols, etc.). After an application profile is modeled, new endpoints can be added without making any changes to the existing policy. After Cisco ACI fabric understands the intended policy, more nodes can be added without the need for any user intervention.

Key value-props of Cisco ACI: Zero-Touch Addition and Removal of Cassandra Nodes

The addition and removal of Apache Cassandra nodes to a cluster demonstrate the power of Cisco ACI application profiles. In an existing environment, you frequently may need to add more Cassandra nodes to a cluster to accommodate growth in the database, increase resiliency, or add computing power. In a non-ACI envt, each node added required manual configuration of the network infrastructure. This process could take a long time, because the needs of application teams cannot readily be translated into traditional network objects such as VLANs, subnets, IP addresses, quality-of-service (QoS) policies, access control lists (ACLs), etc

With Cisco ACI, adding a new node is zero-touch operation. Cisco ACI already has the application profile and components needed to identify the traffic from the new node when it reaches the Cisco ACI fabric, and it can enforce the policy at line rate in hardware immediately. You do not need to notify Cisco ACI of the new nodes.

When Datos RecoverX with Apache Cassandra is run in a Cisco ACI environment, the cluster can elastically expand and contract without the need to manage the details of the network infrastructure underneath. Modern, application-centric policy already is in control and knows how to adapt.

Insieme’s ACI Czar, Carly Stoughton has extensive experience in developing ACI based solutions for hyperscale applications. Carly summarizes the value-props of ACI eloquently as below in her testimonial:

Cisco ACI and Datos RecoverX with Apache Cassandra Validation/Test Environment

We validated the ACI-Cassandra-Datos RecoverX architecture in a production quality Lab environment.

Test Results:

Using CODR architecture that removes media-server bottlenecks, RecoverX was able to complete the initial backup of the entire cluster within 1.6 hours. This result is equivalent to about 60 TB per day of backup performance. And because of the solution’s industry-first semantic deduplication capability, only 1.4 TB of storage capacity was used on the NFS server.

Conclusion:

As more next-generation applications use highly distributed nonrelational databases such as Apache Cassandra and MongoDB, catering to the needs of the application becomes increasingly important from the perspective of the infrastructure. Insieme Business Unit will continue to lead the research and exploration of integrated solutions featuring other web-scale solutions coupled with Cisco ACI.