Cassandra as a Service
Leave the management, monitoring and maintenance of Apache Cassandra and the underlying infrastructure to the experts.
Our customers use our service because they:
- want to focus development resources and efforts on their app and core business,
- need to get it right the first time by using experts in Cassandra,
- have an existing investment in Cassandra that need some help to stabilise their environment,
- are looking to adopt and use Cassandra, but haven’t got time to navigate the new tech curve.
CONTINUOUS AVAILABILITY ARCHITECTURE
DISTRIBUTED CLUSTER NODES
Cassandra allows distribution of cluster servers across cloud provider failure zones. Cassandra’s rack concept is designed to group servers within a cluster where the likelihood of a failure is correlated and to provide ongoing availability despite the failure of a group of servers.
Instaclustr’s architecture maps Cassandra’s rack concept to AWS’s availability zone and Azure’s fault domain concept which designate groups of servers with a correlated likelihood of failure. The servers in a cluster are then distributed equally across racks. This means that the deployed cluster architecture will be optimally aligned with the cloud provider’s likely availability.
RESILIENT TO FAILURE
Cassandra, by design, is resilient to failure of multiple nodes. Additionally, if your environment is configured correctly failure of multiple nodes can still provide 100% availability. With a consistency level of one, Cassandra will still be able to read and write data with the failure of up to two complete rack’s worth of servers.
Instaclustr’s architecture and recommended customer configuration distributes at least one copy of each piece of data in each rack. Once nodes are restored, Cassandra will automatically bring the servers in sync with any missed writes.
Cassandra’s architecture allows nodes to be replaced within a cluster with no interruption to the service provided by the cluster. Instaclustr’s automated provisioning system makes these operations easy and highly reliable. This also applies for the addition of new nodes to support rapid scaling.
Clusters managed by Instaclustr can therefore be scaled out or have faulty nodes replaced (as is often required in a cloud environment) in short time frames with very little risk.
ZERO RELIANCE FOR RUNTIME
Customer nodes are not reliant on Instaclustr software for runtime services. While Instaclustr’s own software is engineered and tested to high standards,we recognise that your application should not depend on our software to continue to run.
At Instaclustr we have engineered our system to ensure there are no runtime dependencies between a customer’s Cassandra service being available and Instaclustr’s management and provisioning systems operating.
“Instaclustr has been instrumental in helping us manage, scale, and optimize our Cassandra deployment, allowing us to focus on delivering a great product to our customers.”Jason Wu
“Instaclustr provided us with a method for getting underway quickly with Cassandra and also delivered the support and expertise necessary to help us make our database as efficient as possible.”Andre
“Instaclustr has enabled us to get underway quickly, the support team have been there from the beginning helping us to get it right the first time with our schema and architecture.”Richard Wilson
OUR MANAGED SERVICE
24/7/365 EXPERT SUPPORT
Customer support delivered by our specialist on-call support engineers with thousands of hours DevOps experience with Apache Cassandra and complementary technologies.
MONITORING & ALERTING
Node monitoring and alerting for events of interest including performance and latency, disk capacity and node responsiveness. Customized alerting is also possible through our monitoring architecture.
ZERO OUTAGE MIGRATIONS
We work with you as part of our service to migrate your data from an existing self-managed environment to the Instaclustr environment with zero outage and continued availability throughout the process.
BACKUPS & SNAPSHOTS
We schedule off-node backups of all data on a daily basis and these are available for restoration in the event of a significant disaster in your cluster. Additionally, Cassandra automatically takes snapshots of your data when you perform certain high-risk operations (such as truncations) and we can assist with restoration where necessary to retrieve your data.
UPDATES & PATCH MANAGEMENT
Our high-availability architecture ensures continuous operations through node upgrades including database version upgrades. We also provide continued monitoring and testing of for patch and security updates and we apply these to your nodes as required. All with zero downtime.
HEALTH CHECKS & REPAIRS
We regularly assess the health of your cluster both through our automated systems and manual checks. Where we detect issues we either undertake repairs, often with our automated operations tooling, or provide advice for you to address them where they relate to your application’s usage of Cassandra.
Customised dashboard with access to graphs and key metrics to monitor the health of your nodes and clusters. We also provide an interface for billing and overall management of your Instaclustr services.
DEDICATED SUPPORT PORTAL
The Instaclustr Support Portal provides a separate and dedicated interface for submitting and tracking support requests and also accessing all of our support and user documentation. Our customers can also access our Instaclustr Status Page through this portal providing indications of incidents or planned outages.
SPEED SPEED SPEED
We have tuned our nodes to perfection and you can expect sub-10ms latency where the application is hosted in the same region as the Cassandra cluster. Tuning and the performance characteristics of Cassandra enable us to achieve this amazing performance.