When choosing the right storage system for workloads original cloud, it is important to consider not only how well the storage system will perform during the deployment, but also how the storage system can be operated. As you evaluate the storage system Kubernetes and how they will perform on Day 2, the production, the following are some issues you need to consider.
1. Monitor performance
Monitoring for both fault and performance problems are one of the most important characteristics for each part of the application, including the storage systems it relies on. It is important for both errors arising debugging and performance tracking. You can browse https://kubevious.io/blog/post/5-useful-monitoring-tools-for-kubernetes to get more information on Kubernetes monitoring.
With proper storage and system monitoring capabilities, you can track changes in performance over time and the number immediately punished decrease in performance, so that the database is not just saying 'slow' you will be able to measure the performance issues.
Capable of monitoring faults and performance degradation is very important, such as instrumentation to gain visibility into how the storage system works and how to connect to the rest of the application. With this in place, there is something wrong when you can quickly find and fix the problem.
The upgrade is an inevitable part of any application lifecycle. You will need a new version of Kubernetes, a version of the new data, the new application version. Kubernetes very effective in orchestrating and managing the upgrade, but the storage system needs to be flexible enough to work with Kubernetes during a rolling upgrade. Kubernetes utilize multiple storage systems Operator, which is a code that automates the deployment and product life cycles for the system.
3. High availability
All applications must assume that something will go wrong at some point there will be a failure, the nodes, and the disk will be lost, the network will fail. Your storage system should be built with high availability in mind to provide redundancy and protect your data in the event of infrastructure failure.