NodeChef Redis Hosting

Nodechef runtime

Deploying Redis to NodeChef bare metal cloud platform is as simple as providing a name for your deployment, select a region and clicking a deploy button.

The NodeChef platform takes care of Redis setup, container orchestration and automatic fail-over to slaves. With a simple web interface, you can deploy, link your App instances in a few clicks.

NodeChef bridges the gap between fortune 100 companies that can afford to run and manage an in-house dedicated infrastructure and small and mid sized businesses that seek dedicated hosting at an affordable price.

NodeChef managed bare metal cloud is more affordable and also outperforms VPS hosting and can save you tons of work configuring and managing your application.

NodeChef Redis deployments allow developers to focus on what they do best (developing applications) and avoid the guesswork of capacity planning. Your deployment autoscales without any downtime to accomodate your workloads. In addition to autoscaling, Redis deployments include built-in backups every 4 hours using the Redis BGSAVE command. Backups can also be archived to your own S3 bucket.


Architecture

Each Redis node in the cluster consists of a proxy exposed to the internet and the Redis Server only accessible from the local host. The proxy is responsible for IP Whitelisting and terminating SSL connections. Each HA Redis deployment also comes with 3 monitoring nodes, that consistently check the state of the cluster.


Layer Role
Proxy Clients connect directly to the proxy. The proxy can accept and terminate SSL connections and forward the request to the Redis Server. The proxy also enforces any IP whitelisting rules.
Redis Server The actual Redis instance listening on local host for connections from the proxy.
Monitor nodes 3 independent nodes responsible for monitoring the state of the Redis Cluster. If the master node fails, one of the monitor nodes initiates the process to promote a slave to the role of a master node. The monitoring nodes are only enabled for HA deployments. If you only deploy a single node without replication, the monitor nodes are not activated.