Hosts, clusters, clouds and logs
Hosts, clusters and clouds, with the logs they write
One collector gathers host metrics, Kubernetes state and logs on Linux, macOS, Windows, Docker and Kubernetes, and cloud discovery maps your AWS, Azure and GCP accounts onto the hosts and clusters you already watch.
- Linux · macOS · Windows
- Docker · Kubernetes
- AWS · Azure · GCP
5 products
Five products for the systems you run
They share one workspace, so a cloud instance links to its host, a host shows its logs, and a service shows the pods it runs on.
Infrastructure monitoring
Every host that reports host metrics appears in a filterable list and a hexagon map, grouped by cloud, region, operating system or tag. Open a host to see CPU, memory, disk, load and network charts, its recent logs and the collector running on it.
- Linux · macOS · Windows
- Host map
Kubernetes and containers
Browse clusters, nodes, namespaces, pods and workloads, with metrics, YAML and events for each, and compare pod usage with requests and limits. Container and image views cover Kubernetes and Docker, and the collector installs into a cluster as a DaemonSet with Helm.
- Helm · DaemonSet
- Kubernetes and Docker
Cloud monitoring
Turn on cloud discovery in a collector policy to list your AWS, Azure and GCP accounts with their instances, databases, load balancers, storage, functions and Kubernetes clusters. Discovered instances and clusters are matched to the hosts, clusters and services you already monitor.
- AWS · Azure · GCP
Log management
Search logs with free text and key:value filters, group them into patterns or transactions, and chart the results. Pipelines in the collector parse JSON, regex, key-value and CSV logs and redact or drop what you do not need, and lifecycle rules move older data to cold storage and delete it after the days you choose.
- JSON · regex · key-value · CSV
- Hot and cold tiers
- Tested pipelines
Metrics explorer
Every metric in the selected time window appears as a card you can break down by a dimension and open in detail, whether it arrived over OTLP, a Prometheus scrape or an integration. For deeper questions, a read-only SQL mode runs ClickHouse queries and can draft them from a plain-language question.
- OTLP · Prometheus scrape
- Read-only SQL
One platform
Connected to the other areas
Infrastructure is the context the other areas link back to.
- ApplicationsA service's Infrastructure tab lists the hosts, pods and images it runs on, and trace samples show the log lines that carry their trace ID.
- NetworkOn Linux hosts and Kubernetes nodes, the collector records connections between services with throughput, retransmits and round-trip time.
- SecurityThe collector inventories the software in running Linux applications, and approved host response actions run through it.
- Operations & AIMonitor templates cover host metrics, Kubernetes and logs, and alerts on the same resource fold into one problem.
Install the collector and see your hosts
Install scripts for Linux, macOS and Windows, a Docker image, or a Helm chart for Kubernetes.