Cases8 cases

3 to 7 of 8

1. IL-0083 Intermittent 5xx errors in a multi-tenant Kubernetes cluster came from internal DNS: the node-local dnsmasq hit its 150 concurrent query limit, did not cache cluster.local answers, and ndots search lookups flooded CoreDNS

   Adevinta2023

   Symptoms

   On a Friday afternoon a tenant of SCHIP, Adevinta's multi-tenant Kubernetes runtime, reported a spike of HTTP 5xx errors in some of its services. The drops in quality of service came back periodically through the following days, each one short, and were reported mostly around midnight.

   High error rateDNS resolution failuresHigh latency

   Root cause

   DNS or network fault

   Component

   CoreDNSClient-side DNS resolver and cacheTelemetry agent

   Contributing factor

   Unreviewed defaultsUnknown limitIgnored early signalUnverified assumption

   Detection

   A tenant reported a spike of HTTP 5xx errors in some of its services; the DNS Latency SLI had already been degraded for some time without being acted on.

   Remediation

   • raised the dnsmasq concurrent query limit from 150 to 1500
   • added a CoreDNS header block that sets the ra and aa flags and clears rd, so dnsmasq caches cluster.local answers
   • configured dnsmasq on each node to answer NXDOMAIN for unresolvable search-domain names instead of forwarding them to CoreDNS
   • changed the fluent-bit log router hostname to a fully qualified name
   • asked the affected tenants to use fully qualified names for internal calls

   Lessons

   The DNS Latency SLI had shown the problem long before anyone looked at it. An SLO only protects you when its SLIs are trusted and a drop is investigated. The investigation took a month partly because several on-call groups worked on it separately, without one place for observations and hypotheses, and sometimes chased hypotheses already discarded. Disprove one hypothesis before starting the next, and collect data the metrics system lacks: DNStop showed the garbage lookups that no dashboard did. With ndots:5, short service names multiply DNS traffic; a node-local cache only helps if it can cache the cluster DNS answers at all.

   View case
2. IL-0105 Replacing the AWS VPC CNI during a Terraform refactoring detached the ALB target groups, and every service in the Kubernetes cluster returned 503

   DFDS2021

   Symptoms

   At 08:29 UTC on 22 June 2021 alerts fired because all services in the Kubernetes cluster were returning 503. No inbound traffic was reaching the cluster. A customer reported the problem at 08:32.

   Service outageHigh error rate

   Trigger

   The pipeline replaced the AWS VPC CNI as part of a refactoring that moved EKS add-ons from kubectl to native Terraform management.

   Root cause

   Configuration change

   Component

   AWS VPC CNICloud load balancerTerraform

   Contributing factor

   Unverified assumptionNo configuration tests

   Detection

   Alerts fired at 08:29 UTC, eight minutes after the VPC CNI was replaced; a customer reported the issue at 08:32.

   Remediation

   • committed a fix that reverted the infrastructure-modules reference
   • re-triggered the pipeline after it failed on an AWS API timeout
   • let the pipeline reattach the target groups to the ALB

   Action items

   • Revert the ALB de-registration delay to the default of 300 seconds
   • Update the roll-over script to set the ALB de-registration delay to zero temporarily before a roll-over
   • Create an ingress test suite

   Lessons

   The change to the de-registration delay had been weighed mainly for performance. Critical pipeline runs should be watched while they run. Everyone should receive the alerts and react to them, and alert messages that were unclear should be reworded.

   View case
3. IL-0193 A GCP persistent disk degradation took down a Cassandra StatefulSet whose OrderedReady policy blocked recovery, and Metrictank could not reconnect to the recreated pods, causing a 23-hour Graphite query outage

   Grafana Labs2019

   Symptoms

   An alert fired because some Metrictank instances could not drain their write queue, which meant they could not write to Cassandra. Google had not yet published an incident, but a team member's Discord server, also hosted on GCP, was down, which pointed at GCP; Google confirmed an SSD persistent disk incident about 90 minutes after it began.

   High error rateNode NotReadyPods PendingService outage

   Trigger

   A Google Cloud incident in which SSD persistent disks performed so poorly that many I/O operations failed for an extended period.

   Root cause

   Cloud provider outage

   Component

   Persistent volumes and stateful setsApplication databaseNode root disk

   Contributing factor

   Overlapping failuresUnpatched known bugUnverified assumption

   Detection

   An alert that Metrictank write instances could not flush their write queue to Cassandra. Error rates alone rose too slowly to notice quickly, because most queries were served from memory.

   Remediation

   • Added compute nodes hoping new nodes could attach the disks, which did not work
   • Deleted and recreated the Cassandra StatefulSet with an updated pod management policy, so all pods but one started
   • Raised Kafka topic retention to avoid data loss during the long downtime
   • Temporarily added nodes so Metrictank restarts were not slowed by CPU starvation
   • Restarted all Metrictank instances, which took more than 20 hours, so they connected to the new Cassandra IPs

   Action items

   • Re-instantiate the gocql connection in Metrictank when it is detected dead, so it reconnects even when all Cassandra IPs change
   • Fix the Metrictank write-queue deadlock by updating an internal data structure with atomics instead of a lock

   Lessons

   An OrderedReady StatefulSet cannot recover past a single pod whose volume will not attach; when every pod is down, the ordering that protects normal start-up blocks the whole cluster. A client that learns peer addresses only from live peers needs a way to go back to the service name when all peers are replaced at once. A long-tolerated bug with a config workaround can surface at the worst time; this one turned out to have a simple fix. Writing incoming data to a Kafka write-ahead log first meant all data could still be ingested after recovery, and Kafka on non-SSD disks escaped the disk incident. With both Metrictank bugs fixed, the team expects a similar event to take less than 3 hours to recover from instead of 23.

   View case
4. IL-0066 Broken ConfigMap change lay dormant for weeks, then took down Pusher's edge proxy when a cluster upgrade replaced its pods

   Pusher2019

   Symptoms

   During a cluster upgrade, all traffic to Pusher's products stopped. The speaker recalls that requests were failing with 500 errors.

   Service outageHigh error rate

   Trigger

   A cluster upgrade replaced nodes, so the bridge pods restarted and loaded a ConfigMap that had been changed three or four weeks earlier.

   Root cause

   Configuration change

   Component

   ConfigMaps and Secrets

   Contributing factor

   Change with a delayed effectUnverified assumption

   Action items

   • Build and run Wave so that changes to ConfigMaps and Secrets roll out new pods immediately

   Lessons

   A ConfigMap change does not restart the pods that use it, so for an application that cannot reload its configuration the change takes effect only when pods are next replaced, possibly weeks later during unrelated work. Tie configuration to the pod template so a bad change fails at the moment it is made.

   View case
5. IL-0064 CPU-heavy pods packed onto one Google Cloud node drove it to kernel panics, and the crash moved from node to node until the site was down

   Moonlight2019

   Symptoms

   On Friday, January 18, Moonlight's API began returning occasional errors that looked like connection problems with its Redis database. The API checks Redis on every authenticated request to validate the session. At the same time the Kubernetes monitoring reported some nodes and pods as unresponsive.

   High error rateIntermittent connection failuresNode NotReadyService outageSustained high node CPU

   Trigger

   The Kubernetes scheduler put several CPU-heavy pods on the same node, including all three replicas of the website and the server-side rendering pods.

   Root cause

   Capacity exhaustion

   Component

   Pod scheduling and placementLinux kernelApplication database

   Contributing factor

   Single point of failureUnverified assumptionAutomation amplification

   Detection

   API errors that looked like Redis connectivity problems, and Kubernetes monitoring that showed nodes and pods as unresponsive. Google Cloud Support later found the link between sustained 100% CPU and kernel panics.

   Remediation

   • added anti-affinity rules to all major deployments so their pods spread across nodes

   Action items

   • Keep working with Google Cloud support to find and fix the cause of the node kernel panics

   Lessons

   Three replicas give no fault tolerance if the scheduler places them all on one node; anti-affinity rules make the spread explicit. A node failure that comes from the workload itself moves with the workload when the scheduler reschedules it.

   View case

By root cause5