Mirantis Kubernetes Engine 3.7.0 delivers ZeroOps for Day 2 operations

As more businesses adopt cloud computing, IT administrators, platform engineers, and application developers alike are struggling with the burden of cloud operations. That’s why Mirantis is strategically focusing on ZeroOps — a radical approach to agile IT that combines automated platforms with managed services to remove the friction of operations throughout the cloud stack. ZeroOps is transformative for organizations because it frees technical staff to focus on more strategic activities that help their business grow. 

The new Mirantis Kubernetes Engine (MKE) 3.7.0 release delivers exciting new ZeroOps features that simplify Day 2 operations, strengthen security compliance, and provide richer observability. Whether you prefer to let Mirantis cloud experts manage your MKE operations, or whether you want to handle it yourself, the new ZeroOps features offer valuable improvements in lifecycle management, certificate authority control, and systems monitoring that will directly benefit many customers.   

ZeroOps upgrades: Automated rollbacks

We always encourage customers to run the newest versions of our platforms, so they can benefit from the latest features and enhancements. Some customers naturally hesitate due to the inherent risk of upgrading production systems. MKE 3.7.0 minimizes this risk through zero-touch automated rollbacks, a crucial component for delivering effortless ZeroOps upgrades. Now if a problem occurs during an upgrade, by default the system quickly reverts to your prior MKE version, so you can keep using your clusters. 

Customers typically need to perform MKE upgrades within a short maintenance window (often over a weekend) that is scheduled weeks or months in advance. While most MKE upgrades successfully complete within the designated time frame, if an upgrade malfunctions and cannot be quickly resolved, there isn’t always enough time left to rebuild the MKE cluster from a backup, leading to potential downtime. Support for automated rollbacks eliminates this issue, giving customers the confidence to upgrade more frequently and improve their overall MKE experience by running the latest code. See Upgrade an MKE installation in the MKE docs.

ZeroOps certificate authority management: Internal CA certificate rotation

Many of our most safety-conscious customers in financial services, defense, law enforcement, and other highly-regulated industries need to tightly control both internal and external authentication certificates for security compliance. While MKE previously enabled customers to rotate external certificates, now MKE 3.7.0 also provides customers the flexibility to rotate internal CA certificates whenever needed. Customers can choose to generate new certificates automatically or provide their own internal cluster root CA certificates. 

Better yet, we’ve also made it easy to rotate internal swarm or MKE cluster root CA certificates, without having to file a support ticket — it just takes a single command! See Manage MKE certificate authorities in the MKE docs.

ZeroOps reporting: Component health monitoring

Continuous, rigorous monitoring is crucial to ensure maximum production uptime of MKE deployments. Now with MKE 3.7.0, operators have more fine-grained observability than ever before, with the ability to track the health of each key component of the platform. This includes etcd, RethinkDB, Calico, as well as Kubernetes-specific components (kube-api-server, kube-controller-manager, kube-proxy, and kubelet), with all metrics exposed through Prometheus. The MKE docs explain all the key metrics to help you understand their meaning, how to assess your system health, and which mitigations to take if problems occur.

Component-level health monitoring is especially important for many of our large financial services customers, as it helps operators to ensure stability, low latency, and high performance. Troubleshooting is faster for issues that may arise from individual platform components. For OpsCare Plus customers, component-level data feeds into the Mirantis incident management system, so support engineers can proactively correct customer environments. 

Besides component-level monitoring, MKE 3.7.0 also gives operators much greater flexibility to configure MKE Prometheus resource limits, use their own Prometheus server, or integrate their preferred corporate monitoring solution. These options enable customers to fine tune observability systems for their specific use case. See MKE component metrics in the MKE docs.

Streamlining container infrastructure operations through ZeroOps

Most ideas for ZeroOps product features, including these features in MKE 3.7.0, originate from real operational problems affecting current customers. There are hundreds of MKE customer deployments in production worldwide, and the Mirantis product team regularly analyzes trends in support incidents to identify which problems have the greatest customer impact and would be most valuable to eliminate through ZeroOps.

To learn more about MKE 3.7.0, please view the release notes.