kloudedocs
Browse documentation
On this page

PRODUCT GUIDE

Monitoring

Inspect cluster and cloud-server resource metrics in Kloude, choose the target and time range, and investigate capacity or availability alongside your application monitoring.

Monitoring surfaces health and resource signals for Kubernetes clusters and cloud servers without leaving the portal. It is part of day-to-day operations on Kloude.

Open Monitoring from the portal navigation when you need a quick capacity or health check. Pair it with the resource views inside Clusters (Pods, Events, Cloud Shell) and with your own APM or log stack for application depth.

What you see in the panel

Select the cluster or server you want to inspect in Metrics. Use the available filters to choose the time range, refresh interval, and view. The charts and metrics depend on the selected resource; cluster and server metrics come from different monitoring sources.

The cluster Overview includes resource gauges such as:

  • Disk — root filesystem usage
  • Memory — cluster memory usage against total capacity
  • CPU — processor load against total cores

Sample numbers in the tutorial screenshot are illustrative only; always trust the live values for your own cluster.

When to use it

Check CPU, memory, and workload status after you deploy an app; look at capacity before you scale a node pool; and confirm a cluster is healthy after create or upgrade.

Tutorial: check health in the portal

Open Monitoring

Select Monitoring in the portal navigation.

Stay on Metrics

Keep Metrics active to inspect the selected resource. Open Billing separately when investigating resource cost.

Pick the resource and filters

Select the right cluster or server. Set the time range and refresh interval so they match the change you just made (for example the last few hours after a deploy).

Read Overview

In Overview, review the Disk, Memory, and CPU gauges. A sudden memory or CPU spike can explain Pending pods or resource pressure; treat a filling root disk as urgent.

Act in Clusters

Scale pools or inspect workloads from Clusters. Open Support if the platform itself looks stuck.

Monitor a cloud server

Choose a cloud server as the monitoring target, or follow the Monitoring link from that server’s details. Check CPU, memory, and other available signals over the time window relevant to your question. Server metrics and cluster metrics are collected separately, so available views can differ.

If a server cannot be reached, combine the metrics with its lifecycle state, networking, firewall rules, and browser console. Record the resource name and time range if you need to open a support ticket.