Cases12 cases

6 to 10 of 12

1. IL-0046 Prometheus restarted again and again because its WAL replay took longer than the startup probe allowed

   UK Ministry of Justice2024

   Symptoms

   At 12:32 Prometheus on the Cloud Platform was not available after a planned change, and the Prometheus pod had restarted several times. The platform lost its monitoring and had missing metrics for the period.

   Monitoring outageProbe failures

   Trigger

   A planned change restarted Prometheus at a time when the WAL count was high.

   Root cause

   Capacity exhaustion

   Component

   Prometheus

   Contributing factor

   Unreviewed defaultsUnknown limit

   Detection

   The team saw that the Prometheus pod had restarted several times after the planned change.

   Remediation

   • raised the Prometheus startupProbe threshold to 30 minutes

   Action items

   • Perform planned Prometheus restarts when the WAL count is lower
   • Set the default CPU and memory requests to meet maximum usage
   • Create a test setup that recreates the live WAL count
   • Explore the memory-snapshot-on-shutdown and auto-gomaxprocs feature flags
   • Explore storing WAL files remotely in a different location
   • Look into a blue-green Prometheus to test changes on a live-like setup before applying them to live
   • Spike into Amazon Managed Prometheus

   View case
2. IL-0194 After a kube-apiserver restart, pod LIST requests from every pod of a Vector DaemonSet nearly tripled etcd memory and restarted all master nodes

   Palark2023

   Symptoms

   One evening, right after an ordinary restart of kube-apiserver in a customer's production cluster, etcd memory consumption rose almost threefold. The sudden growth led to a cascading restart of all the master nodes.

   Control plane unresponsiveDatabase overload

   Trigger

   An ordinary restart of kube-apiserver, after which every pod of the log collector DaemonSet sent its list requests at the same time.

   Root cause

   Capacity exhaustion

   Component

   etcdKubernetes API serverTelemetry agent

   Contributing factor

   Unknown limitUnreviewed defaults

   Detection

   Memory graphs showed etcd memory on the master nodes almost tripling after a kube-apiserver restart; the audit logs then pointed to one DaemonSet's list requests.

   Remediation

   • created a PriorityLevelConfiguration named limit-list-custom that queues requests (assuredConcurrencyShares 5, 16 queues, handSize 4, queueLengthLimit 50)
   • created a FlowSchema that routes list and get requests on pods from the log collector's service account to that priority level

   Action items

   • Submit a pull request adding resource version support to the Kubernetes Rust library
   • Open a pull request to the Vector log collector to address the startup list requests
   • Apply the same flow-control approach to a similar problem found with the Cilium CNI

   Lessons

   A field selector reduces what kube-apiserver returns, not what it reads from etcd: listing the pods of one node still reads every pod in the cluster. A DaemonSet whose pods all list at startup multiplies that by the node count each time kube-apiserver restarts. Where exact freshness is not needed, resourceVersion=0 lets kube-apiserver answer from its cache; where the client cannot set it, API Priority and Fairness can queue the requests so they reach etcd gradually.

   View case
3. IL-0013 The conntrack table filled up on downsized GKE nodes and a high-traffic aggregation proxy saw climbing connection resets

   loveholidays2020

   Symptoms

   Shortly after a round of changes, loveholidays saw connectivity problems and raised error rates on its aggregation service. The service is a proxy between two other services of more than 100 instances each, handles over 10,000 requests per second continuously and can spike to four times that.

   Connection resetsHigh error rateIntermittent connection failures

   Trigger

   Three changes together: a change in service architecture, halving the RAM of the nodes that run the aggregation service, and sharding a downstream service, which doubled the number of connections to track.

   Root cause

   Capacity exhaustion

   Component

   conntrack tablekube-proxyNode poolHAProxy

   Contributing factor

   Unknown limitNo capacity planningNo alert on limit

   Detection

   Rising error rates and climbing connection reset errors in HAProxy, then node_exporter conntrack metrics compared with their limit.

   Remediation

   • moved the aggregation service to nodes with more RAM
   • limited the number of aggregation service pods per node

   Action items

   • Alert when conntrack table use goes above 75%, with (node_nf_conntrack_entries / on (pod) node_nf_conntrack_entries_limit / on (pod) group_right kube_pod_info) > 0.75

   Lessons

   The conntrack limit is tied to node memory, so resizing a node pool silently resizes the connection table, and on GKE the ratio did not match the commonly cited formula. A high-traffic proxy on small nodes can exhaust it, and the symptoms are connection resets and timeouts, so conntrack use needs its own metric and alert.

   View case
4. IL-0015 A GKE cluster's /16 pod address range ran out at 256 nodes, leaving new pods Pending and node scale-ups failing

   loveholidays2020

   Symptoms

   A colleague wrote in Slack that his application was taking a long time to deploy. A deployment that normally finishes in minutes had been running for about an hour: about half of the new pods were running and serving traffic, and the other half were Pending.

   Pods PendingIP address exhaustionScale-up failure

   Trigger

   The cluster reaching 256 nodes, the most its /16 pod secondary range could hold at 256 addresses per node; every further node creation failed.

   Root cause

   Capacity exhaustion

   Component

   VPC subnet or pod address rangeCluster autoscalerNode pool

   Contributing factor

   Unreviewed defaultsUnknown limit

   Detection

   A colleague reported in Slack that a deployment was slow; the pending pods' events and the cluster autoscaler status ConfigMap showed the exhausted IP space.

   Remediation

   • lowered the maximum pods per node on selected node pools, cutting IP consumption by about 30%
   • moved some node pools to larger instance types

   Action items

   • Lower the maximum pods per node on node pools that do not run many pods
   • Use larger instance types in some node pools

   Lessons

   In a VPC-native GKE cluster the number of nodes, not the number of pods, uses up the pod range: by default each node reserves 256 addresses, so even a /16 is full at 256 nodes. Maximum pods per node and node size are the practical levers; larger nodes pack better but take more pods down when they fail.

   View case
5. IL-0037 Image registry rate-limited a whole region after a crash-looping DaemonSet with imagePullPolicy Always pulled through three NAT instances

   Datadog2019

   Symptoms

   Users reported that their jobs were not starting, with some sign that image pulls were involved. Many pods across the cluster were in error states related to image pulls, while applications that were already running were not affected.

   Image pull failuresService outage

   Trigger

   A change to the permissions on a storage bucket that one application needed to read. The application could no longer start, and its DaemonSet, running on roughly a thousand nodes of the cluster, went into a crash loop with imagePullPolicy: Always, so every restart pulled the image again although it was already on the node.

   Root cause

   Capacity exhaustion

   Component

   Container registry and its credentialsNAT gatewayApplication databaseAdmission webhook

   Contributing factor

   Automation amplificationSingle point of failureUnknown limitHarmful manual actionExcessive blast radius

   Detection

   Users reported that their jobs were not starting; monitoring had not caught it first.

   Remediation

   • replaced the affected NAT instances so registry traffic came from new source IPs
   • reconfigured the static IPs and firewall rules so applications could reach the Cloud SQL database again

   Action items

   • Add an admission webhook that rejects floating image tags, starting with latest
   • Apply the image tag webhook to workload objects only, not to pods, after it blocked pods created by in-cluster controllers

   Lessons

   The speakers' takeaway is to be very careful with DaemonSets in large clusters: they say they had broken everything DaemonSets depend on, including DNS, Vault and the image repository, and that it is very easy to break things when managing them by the thousand. imagePullPolicy: Always made every crash-loop restart pull the image again. Sending all registry traffic through three NAT instances put the whole region behind three source IPs, which the registry rate-limited, and changing those IPs cut the connections to a critical database reached over the internet through them. An admission webhook on pods rejects the pods that controllers in the cluster create, where no user sees the error.

   View case

By root cause1