Logit.io
Solutions

Kubernetes Monitoring

Watch for Kubernetes errors and be notified. Logging, metrics, and monitoring for Kubernetes clusters with Logit.io.

observability stack
Ingest
Correlate
Visualize
Act
Logs
Logstash · OpenSearch
12.4k/s
Metrics
Prometheus · Grafana
10k series
APM
OTLP · traces
p95 84ms
--:--:-- INFO request.completed duration_ms=42 route="/api/v1/orders"
--:--:-- WARN latency.spike service=checkout p95=820ms threshold=500ms
--:--:-- INFO trace.exported spans=128 backend=jaeger status="ok"
--:--:-- INFO metric.scrape target=prometheus job=k8s-pods samples=8421
--:--:-- INFO log.shipped bytes=184032 index=logs-prod
--:--:-- WARN auth.failure ip=203.0.113.42 attempts=3 action=rate_limit
--:--:-- INFO alert.routed severity=high channel="#incidents" dedupe=on
--:--:-- INFO dashboard.refresh uid=ops-overview panels=14 cache=hit
--:--:-- ERROR disk.pressure node=worker-3 usage=92% reclaim=started
--:--:-- INFO pipeline.batch size=2048 lag_ms=18 status=healthy
--:--:-- WARN queue.backpressure topic=ingest depth=1200
--:--:-- INFO otel.export endpoint=collector.svc spans_ok=512
--:--:-- INFO search.query hits=1284 took_ms=37 index=logs-*
--:--:-- INFO retention.policy applied hot=14d warm=30d
--:--:-- WARN tls.cert.expiring host=ingest.logit.io days=12
--:--:-- INFO ha.failover check region=eu-west status=ready
--:--:-- INFO request.completed duration_ms=42 route="/api/v1/orders"
--:--:-- WARN latency.spike service=checkout p95=820ms threshold=500ms
--:--:-- INFO trace.exported spans=128 backend=jaeger status="ok"
--:--:-- INFO metric.scrape target=prometheus job=k8s-pods samples=8421
--:--:-- INFO log.shipped bytes=184032 index=logs-prod
--:--:-- WARN auth.failure ip=203.0.113.42 attempts=3 action=rate_limit
--:--:-- INFO alert.routed severity=high channel="#incidents" dedupe=on
--:--:-- INFO dashboard.refresh uid=ops-overview panels=14 cache=hit
--:--:-- ERROR disk.pressure node=worker-3 usage=92% reclaim=started
--:--:-- INFO pipeline.batch size=2048 lag_ms=18 status=healthy
--:--:-- WARN queue.backpressure topic=ingest depth=1200
--:--:-- INFO otel.export endpoint=collector.svc spans_ok=512
--:--:-- INFO search.query hits=1284 took_ms=37 index=logs-*
--:--:-- INFO retention.policy applied hot=14d warm=30d
--:--:-- WARN tls.cert.expiring host=ingest.logit.io days=12
--:--:-- INFO ha.failover check region=eu-west status=ready
+ one managed platform
+ peak overage protection
! bundle up to 30% off

Trusted by engineering teams worldwide

Maersk
Murphy
Ringier
GDS
Guesty
HackerRank
Equal Experts
DevEx
Digitale Medier
xneelo
CAA
Pivotal
Robomed Network
Neoway
Gomo Learning
Department for BEIS
IBM
Broad Institute
The Honest Company
Traels
De Banke
Dofinity
BioCatch
Kainos
Youredi
Flux Music
Goji
Ving
HypSports
Boston Logic
Double Jump

The move to containerized applications and microservices architecture has transformed the way we develop and deploy applications. It is crucial to monitor the performance of these containers and ensure that they are available 24/7 and that there are no security or performance issues.

Kubernetes is the tool of choice for many DevOps to run containers on clusters of virtual machines and many enterprises are opting for Kubernetes-as-a-Service (KaaS) models to deploy and manage Kubernetes clusters.

A variety of tools can be used for monitoring Kubernetes environments, including monitoring solutions which are native to Kubernetes. However, these tools often lack interoperability with other monitoring systems, making it difficult to truly understand the health of an entire enterprise application environment. Many of these tools often do not provide user-friendly interfaces or dashboards to help users make sense of what's happening within their clusters. By using Logit.io you can get started with monitoring Kubernetes, alongside Docker and a wide variety of additional data sources.

Why should you Monitor Kubernetes?

Kubernetes is the most popular container orchestration platform for deploying, scaling and operating containerised applications. Kubernetes provides APIs for scheduling containers onto nodes in a cluster, replication controllers to ensure that a specified number of copies of a container are running, service discovery to find services inside the cluster, and health checking to monitor the availability of pods.

One way to ensure the expected performance of containers in Kubernetes is with a monitoring platform. Kubernetes monitoring helps your development teams and operations teams effectively manage your cluster.

Logit.io's Kubernetes monitoring solution is built for monitoring both production and development clusters, as they monitor the status of containers, nodes, and network resources. The Logit.io platform can provide a wealth of information about the pods currently running and nodes of the cluster, as well as send error alerts.

pipeline
Ingest
Parse
Index
Query
+streams normalized · schema applied
output ready for search, alerts, and dashboards

Key Kubernetes Metrics to Monitor

Visualising Kubernetes metrics is an excellent way to monitor and manage your clusters, nodes, pods, and applications. Here are some of the most useful Kubernetes metrics that you can visualise in Logit.io hosted Kibana reporting dashboards.

Node Resources: this metric monitors the readiness, disk or processor overload, memory, memory requests, memory limits, allocatable memory, memory utilization, CPU requests and network availability of deployed clusters.

Cluster state metrics: this metric monitors Node status, desired pods, current pods available and unavailable pods and their health.

Kubernetes events: Monitor how pod creation, deletion, starting, and halting affects the performance of your infrastructure by collecting events from Kubernetes.

Control plane metrics: The control plane is Kubernetes' brain and heart. The cluster's optimal operation and efficiency rely on all of its components. Monitoring the state of the Kubernetes control plane is just as critical as monitoring the nodes or apps operating inside.

stream
--:--:-- INFO request.completed duration_ms=42 route="/api/v1/orders"
--:--:-- WARN latency.spike service=checkout p95=820ms threshold=500ms
--:--:-- INFO trace.exported spans=128 backend=jaeger status="ok"
--:--:-- INFO metric.scrape target=prometheus job=k8s-pods samples=8421
--:--:-- INFO log.shipped bytes=184032 index=logs-prod
--:--:-- WARN auth.failure ip=203.0.113.42 attempts=3 action=rate_limit
--:--:-- INFO alert.routed severity=high channel="#incidents" dedupe=on
--:--:-- INFO dashboard.refresh uid=ops-overview panels=14 cache=hit
--:--:-- ERROR disk.pressure node=worker-3 usage=92% reclaim=started
--:--:-- INFO pipeline.batch size=2048 lag_ms=18 status=healthy
--:--:-- WARN queue.backpressure topic=ingest depth=1200
--:--:-- INFO otel.export endpoint=collector.svc spans_ok=512
--:--:-- INFO search.query hits=1284 took_ms=37 index=logs-*
--:--:-- INFO retention.policy applied hot=14d warm=30d
--:--:-- WARN tls.cert.expiring host=ingest.logit.io days=12
--:--:-- INFO ha.failover check region=eu-west status=ready
--:--:-- INFO request.completed duration_ms=42 route="/api/v1/orders"
--:--:-- WARN latency.spike service=checkout p95=820ms threshold=500ms
--:--:-- INFO trace.exported spans=128 backend=jaeger status="ok"
--:--:-- INFO metric.scrape target=prometheus job=k8s-pods samples=8421
--:--:-- INFO log.shipped bytes=184032 index=logs-prod
--:--:-- WARN auth.failure ip=203.0.113.42 attempts=3 action=rate_limit
--:--:-- INFO alert.routed severity=high channel="#incidents" dedupe=on
--:--:-- INFO dashboard.refresh uid=ops-overview panels=14 cache=hit
--:--:-- ERROR disk.pressure node=worker-3 usage=92% reclaim=started
--:--:-- INFO pipeline.batch size=2048 lag_ms=18 status=healthy
--:--:-- WARN queue.backpressure topic=ingest depth=1200
--:--:-- INFO otel.export endpoint=collector.svc spans_ok=512
--:--:-- INFO search.query hits=1284 took_ms=37 index=logs-*
--:--:-- INFO retention.policy applied hot=14d warm=30d
--:--:-- WARN tls.cert.expiring host=ingest.logit.io days=12
--:--:-- INFO ha.failover check region=eu-west status=ready

Real-Time View of your Cluster's Health and Performance

Monitor Kubernetes clusters in real-time and at cluster-wide scale with Logit.io. Monitor the master and all the worker nodes in a cluster and report on processes (Pods, Engines, and Nodes) in each node by using Kibana for visualisation.

Keep track of CPU and memory allocation and utilization, and see your entire cluster at a glance

Use Logit.io's Kubernetes Monitoring Dashboard to help you keep track of your namespaces, pods, services and deployments and display the status of your cluster and the health of every namespace associated with it.

Visualize, alert on, and diagnose the entire state of your Kubernetes cluster in one place. Whether you have 1 node or 1000, the Logit.io platform gives you deep visibility into the health and performance of your Kubernetes infrastructure.

shell
$
logit trace export --protocol=otlp --backend=jaeger
→ distributed spans indexed · p99=84ms

Automated Monitoring

Track pods, nodes, container resource allocations and utilization, readiness, scheduling, and more with Logit.io.

Monitor clusters and services to keep tabs on their health and performance. Spot problems before they affect users or your business: for each pod in a namespace, track its readiness and schedule, CPU and memory allocation and utilization, visualise volume usage, and more.

Take advantage of rich and actionable analytics and visualizations to get a complete picture of your cluster's resource usage and determine how much new workload you can add without overloading your nodes.

Monitor the status of the application services linked with each namespace, as well as the age of the service type and the pods that are connected to each service.

Get tighter control of your deployments and be aware of how they're doing. Get alerted to failed jobs, failed pods, and the longest-running jobs. Run all the metrics you need on managed or unmanaged clusters.

metrics
latency
throughput
errors
p95 latency 84ms · ingest 12.4k/s · error rate 0.08%

Plan. Deploy. Scale. Visualise.

Taking action often starts with gaining better visibility, and Logit.io gives visibility across your entire infrastructure. Logit.io gives users the power to parse and analyse data from their Kubernetes clusters and add alerting rules based on meta-objects such as Deployments, Pods, and other elements.

Easily onboard and collect metrics from services running on Kubernetes-provisioned containers.

If you wish to go beyond visualising Kubernetes data and wish to explore how trace analytics can assist your application development then why not consider using Logit.io's APM solution.

Read more about Application Performance Monitoring (APM)
alerts
1
Detect
2
Enrich
3
Route
4
Notify
! anomaly detected · checkout p95 > 500ms
→ context attached · service map · recent deploy
→ routed to #incidents · ack in 12s

Logit.io For OpenTelemetry

With OpenTelemetry (OTel) and Logit.io, you can gain end-to-end observability effortlessly. Essentially, OpenTelemetry standardises all data from any source, enabling automated analysis and as a result, saves engineers hours every month.

With Logit.io for OpenTelemetry, your data is completely centralized, unified and secure, with this compliance and production-ready distribution of the OpenTelemetry project.

All major observability vendors must support the OpenTelemetry protocol going forward, therefore choosing an analysis service that already supports OpenTelemetry will future-proof your operations.

Find out more about OTel
stack
OpenSearchPrometheusGrafanaJaegerLogstash
$ logit stack status --managed
all services healthy · HA enabled · backups current

Companies Feel The Difference When They Use Logit.io

Internally, Logit.io has made it easier for us to provide better support for our customers, since finding individual messages based on various data in the payload has become easier.

At Youredi, pretty much everyone from our technical support teams through to our professional services teams uses Logit.io.

Youredi

Mats von Weissenberg

CTO @ Youredi

Start your 14-day free trial

No credit card required. Managed OpenSearch, Prometheus, and Grafana from $25/mo.