Cases43 cases

36 to 40 of 43

1. IL-0008 An unresponsive OPA validating webhook crash-looped the API server during a GKE master upgrade, and node auto-repair then recreated every node

   Jetstack2019

   Symptoms

   Jetstack was upgrading the master of a development cluster that many teams use to test their apps during the working day, a regional Google Kubernetes Engine (GKE) cluster in europe-west1. The upgrade ran through Jetstack's GKE Terraform pipeline and did not finish before the 20-minute Terraform timeout, while the GKE console still showed the cluster as upgrading.

   Upgrade stallControl plane unresponsiveNode recreation loopService outage

   Trigger

   A planned master upgrade of the shared development cluster. The same upgrade had been done on another pre-production environment earlier that day, and no API downtime was expected.

   Root cause

   Configuration change

   Component

   Admission webhookGKE control planeKubernetes API serverkubelet

   Contributing factor

   Excessive blast radiusNo health probesMonitoring blind spotAutomation amplification

   Detection

   The first sign was the master upgrade missing the 20-minute Terraform timeout; re-running the pipeline returned the kube-apiserver unhealthy error. The cause was identified with the help of Google Support.

   Remediation

   • Deleted the OPA ValidatingAdmissionWebhook resource during a window of API server access

   Action items

   • Limit the OPA webhook to the namespaces where the policy applies and development teams have access
   • Enable the webhook only for Ingress and Service resources, the only resources the policy validates
   • Add a liveness probe so OPA is restarted when it stops responding, and update the OPA documentation
   • Follow the open upstream Kubernetes issue about this behaviour

   Lessons

   With alerting on API server response times, the team might have seen CREATE and UPDATE requests slow down across the board as soon as the OPA-backed webhook was first deployed. Every workload needs probes: OPA looked so simple to deploy that the team skipped the Helm chart, which adds a livenessProbe for the admission controller containers among other adjustments to the tutorial configuration. Others had hit the same problem, and an upstream issue to improve it was still open.

   View case
2. IL-0052 A cryptocurrency miner ran on every node of the development and staging clusters through a public, unauthenticated Weave Scope dashboard

   JW Player2019

   Symptoms

   Datadog alerted the DevOps team to a high normalised load average, first on the staging cluster (Day 0, 21:06 EST) and the next day on the development cluster (Day 1, 16:53 EST). The load was first put down to a legitimate service that kept scaling up and down in both environments, and because of the volume of identical alerts they were muted until things settled.

   Sustained high node CPUUnauthorised access or a foreign workload

   Trigger

   An automated crawler found the Weave Scope dashboard on a public load balancer that required no authentication. Weave Scope can open a shell in running containers, and its scope-agent containers ran privileged, as root and with the host's root filesystem mounted, so a command started there could run on the host.

   Root cause

   Configuration change

   Component

   Cluster dashboardCloud load balancer

   Contributing factor

   Ignored early signalUnreviewed defaultsExcessive blast radiusMonitoring blind spot

   Detection

   Datadog alerts on high normalised load average, muted at first as noise from a legitimate service; on Day 3 an engineer ran top on a node and found the gcc process at 100% CPU.

   Remediation

   • stopped Weave Scope and removed its deployment from all Kubernetes clusters
   • checked access logs for unauthorised access to the instances
   • confirmed the miner's outbound connections went only to a mining pool
   • isolated one affected node for later investigation
   • rotated out every node in each cluster so all affected hosts were destroyed and rebuilt

   Action items

   • Audit all Kubernetes and deployment YAML files and security groups for services that are unintentionally public or misconfigured
   • Replace load-average alerts with more meaningful monitors and tune them to signal a real anomaly
   • Research tools such as falcon or sysdig for behavioural monitoring and anomaly and intrusion detection
   • Evaluate Istio and Linkerd for observing and controlling end-to-end network traffic
   • Analyse the cost of scanning Docker images and containers for known vulnerabilities
   • Set aside time in architecture discussions to revisit older choices such as running containers as root
   • Share knowledge of large projects, such as Kubernetes version migrations, across the whole team

   Lessons

   Load average was not a useful signal for spotting a problem in the clusters, and muted noisy alerts nearly hid the miner. A load balancer that Kubernetes creates belongs to Kubernetes: a manual security group edit in the AWS console is reverted, so internal exposure must be declared in the Kubernetes YAML. A dashboard that can open shells, in front of privileged agents that mount the host's root filesystem, gives whoever reaches it root on every node.

   View case
3. IL-0035 Worker nodes went NotReady after a new ElastAlert rule without memory limits used up node memory

   NU.nl2019

   Symptoms

   The NU.nl development team got reports that several services on one of its Kubernetes clusters behaved erratically: occasional error pages, slow responses and timeouts. The affected services were mostly internal tools and dashboards.

   Node NotReadyHigh error rateHigh latencyIntermittent timeoutsMonitoring outageOOMKilled

   Trigger

   A new ElastAlert rule, deployed shortly before from the command line, that fired if the Cloudfront-* Elasticsearch indexes stopped receiving documents; those indexes take a few million requests per hour, and the rule made ElastAlert's memory grow without bound.

   Root cause

   Configuration change

   Component

   Container memory limitkubeletPrometheus

   Contributing factor

   No resource requests or limitsUnreviewed defaultsNo alert on limitMonitoring blind spotExcessive blast radius

   Detection

   Users reported error pages, slow responses and timeouts from several services; no alert fired.

   Remediation

   • terminated the NotReady nodes one by one so the auto-scaling group replaced them (the new nodes also went NotReady, but Prometheus and Grafana landed on a stable node)
   • found the ElastAlert pods' memory growth on the Grafana pod memory dashboard
   • fixed the ElastAlert problem that caused the memory growth

   Action items

   • Define alerts on resource usage such as CPU, memory and disk space
   • Keep researching an alerting strategy that combines Elasticsearch, Prometheus Alertmanager and New Relic
   • Consider exporting metrics out of the cluster and running Grafana outside it, on EC2 or a separate Kubernetes cluster
   • Either visualise the important metrics in the ELK stack or make Prometheus and Grafana more available, and improve the metrics strategy
   • Set resource limits for ElastAlert through the Helm values
   • Configure namespaces with default memory requests and limits
   • Isolate customer-facing workloads from build, logging and monitoring workloads (a dedicated node group, or later a separate tools cluster)
   • Deploy everything through automation such as Jenkins pipelines, and consider a GitOps approach
   • Deploy through a pipeline that includes a smoke test step
   • Make Kubernetes and cloud infrastructure work part of the team's sprints, picked up by all team members while pairing with more experienced members, and hold workshops that go deep into specific topics

   Lessons

   Fixing the root cause covers only a small part of what an incident can teach; the chain of smaller failures around it is where most of the learning is. A pod without a memory limit can take whole nodes down, so limits need defaults at the namespace level. Monitoring that runs inside the cluster it watches goes blind with it. Changes installed by hand from the command line leave the rest of the team without a record of what changed.

   View case
4. IL-0075 Kubernetes API servers pegged on CPU when the cluster grew to about 290 nodes, traced to Datadog agents polling the API

   Civis Analytics2019

   Symptoms

   After Civis released its new script infrastructure on Kubernetes internally in September 2018, developers reported that kubectl commands were slow, and users intermittently could not start notebooks and services.

   Control plane lagControl plane unresponsiveHigh latencyHigh error rateDNS resolution failures

   Trigger

   Migrating clients' Python, R and container scripts from the in-house Bocce system to Kubernetes (the project called Croquet) pushed the cluster to about 290 EC2 instances under a large script load.

   Root cause

   Capacity exhaustion

   Component

   Kubernetes API serverTelemetry agentetcd

   Contributing factor

   Unknown limitUnverified assumptionMonitoring blind spotNo capacity planning

   Detection

   Developer reports of slow kubectl commands and user reports of notebooks and services failing to start, then high CPU on the master nodes.

   Remediation

   • moved internal users back to the old Bocce infrastructure
   • resized the three master nodes from m4.large to m4.4xlarge
   • changed the Datadog setup so per-node agents no longer polled the API server for service metadata
   • raised max-requests-inflight for non-mutating requests from 400 to 1200
   • grew the etcd EBS volumes from 20 to 100 GB (100 to 300 IOPS)
   • switched client scripts from the ClusterFirst DNS policy to Default

   Action items

   • Upgrade etcd from 2.2.1 to 3.x once Kops supports a safe migration
   • Solve the in-cluster DNS timeouts so client scripts can talk to other workloads in the cluster

   Lessons

   Kubernetes can reach thousands of nodes only after tuning, and third-party DaemonSets are a common cause of control plane load: an agent on every node that polls or watches the API server multiplies with the cluster size. Datadog, kube2iam and the fluentd metadata filter all showed this. A small configuration change often fixes it, but finding it took load testing and the API server access logs.

   View case
5. IL-0001 Cluster-wide DNS outage after CoreDNS pods were OOM-killed again and again under a burst of DNS queries

   Zalando2019

   Symptoms

   Customers of the Zalando Fashion Store got a high rate of errors on all web product pages and outfit pages for more than an hour. The errors came from the main data aggregation layer service, which runs in one of Zalando's Kubernetes clusters.

   High error rateDNS resolution failuresOOMKilledMonitoring outage

   Trigger

   Calls from the aggregation layer to a single, not very important downstream service started timing out. The aggregation layer returned 404s for them, clients retried, and the resulting spike in requests became a spike in DNS queries to CoreDNS.

   Root cause

   Capacity exhaustion

   Component

   CoreDNSContainer memory limitClient-side DNS resolver and cache

   Contributing factor

   Unknown limitNo cachingRetry stormUnreviewed defaultsExcessive blast radiusMonitoring blind spot

   Detection

   The on-call engineer for the aggregation layer phoned the on-call Kubernetes engineer. Automatic paging did not happen because the cluster's monitoring and alerting ran inside the cluster and failed along with DNS. The Kubernetes engineer then saw that the CoreDNS pods were being OOM-killed.

   Remediation

   • Raised the CoreDNS memory requests and limits by hand from 100Mi to 2000Mi

   Action items

   • Introduce a more resilient DNS setup: CoreDNS fronted by dnsmasq on every node
   • Improve high-level alerting from outside the clusters, so that a cluster problem is known before or at the same time as applications are affected
   • Improve incident response processes and tooling so that any on-call engineer can escalate directly to other teams when automatic paging fails, for example through a chat bot

   Lessons

   The Kubernetes DNS setup was far from resilient enough: a single application could take down DNS for everything in the cluster. The blast radius of one application sending more DNS queries should be limited to one node at most, ideally to single pods. Known Kubernetes DNS problems, such as Linux kernel races in conntrack handling and the ndots 5 problem, were familiar to the team and partly mitigated, but their combined effect on the DNS infrastructure was worse than the team had understood. Monitoring that runs inside the cluster was not robust enough: when the cluster's DNS broke, it could not page anyone, and recovery took longer than it should have.

   View case

By root cause9