Cases13 cases

3 to 7 of 13

1. IL-0499 Discord voice and video calls hung at Awaiting Endpoint for three hours after 17% of sessions stopped at once

   Discord2026

   Symptoms

   From 12:13 to 15:30 PDT on 25 March 2026, voice and video on Discord were heavily degraded. Users mostly could not start or join calls and saw "Awaiting Endpoint" in their call status.

   Service outage

   Trigger

   A routine change to tune the sessions service cluster on Kubernetes gave each pod more CPU and memory and lowered the pod count in proportion. When it was deployed to the first zone at 12:13 PDT, Kubernetes terminated 50% of that zone's pods, and the termination grace period ran out before the sessions could be handed off.

   Root cause

   Configuration change

   Component

   Pod terminationCoordination and key-value storeDatabase proxy and connection pooler

   Contributing factor

   Retry stormUnknown limit

   Remediation

   • Set low syncer spawn rate limits in the calls, streams and guilds services and restarted the voice syncers cluster
   • Restarted voice syncers instances one at a time, and four of the first five recovered fully
   • Brought 15 new voice syncers instances online, doubling capacity and halving the syncers per instance
   • Restarted the last backed-up voice syncers instance, leaving the doubled cluster fully healthy

   Action items

   • Add a validating admission webhook that rejects a scale-down of Elixir workloads until each pod has handed off its processes
   • Replace the Holster.Pool supervisor with a PartitionSupervisor so outgoing connections are created and checked out concurrently
   • Move gun connection process lifecycle management into Holster.Pool so the gun connection supervisor is no longer used
   • Tune the upstream rate limits that call into voice syncers, with load shedding tooling that can force them
   • Add rate limits on endpoint selection in voice syncers to limit outbound connection creation
   • Monitor the HTTP connection pooling mailboxes
   • Review and expand instrumentation and logging for service discovery announcement and voice syncer RPCs to SFUs
   • Monitor the most critical processes across all services instead of aggregate metrics such as overall mailbox health
   • Bring more workloads into Kubernetes so capacity can be scaled with a single command instead of provisioning VMs by hand
   • Re-architect the A/V system to scale its control plane services horizontally across multiple regions

   Lessons

   Discord says a large enough traffic spike will find a bottleneck somewhere; in earlier large-load incidents its ScyllaDB databases limited recovery, and this time the bottleneck moved to the voice services. Facing a capacity constraint, the team asks whether to increase supply, such as adding services, or reduce demand, such as rate limiting inbound traffic. During an incident it matters to know the toolbox, and Elixir and the BEAM give engineers many introspection options. Any slowdown in the hot loop of a single-threaded process can degrade both the service and the user experience.

   View case
2. IL-0304 Intercom's US region was fully down for 71 minutes when its Vitess VTGates dropped all routes to the database shards

   Intercom2026

   Symptoms

   From 19:39 to 20:50 UTC on January 9, 2026, all Intercom services in the US region were unavailable to customers hosted there. Queries from the Intercom application could not reach the database shards, and Inbox Heartbeat Anomaly alarms fired in the US region at 19:40 UTC.

   Service outage

   Trigger

   At 18:56 UTC a configuration in the database management pipeline treated a 'branch unfreeze' action as a signal to start a Vitess v22 upgrade planned for the following week. At 19:30 UTC a MoveTables workflow failed during transient networking issues in the provider's Kubernetes cluster, and the v22 rollback logic applied an empty VSchema.

   Root cause

   Breaking change in an upgrade

   Component

   DatabaseDatabase proxy and connection pooler

   Contributing factor

   Unpinned versionOverlapping failuresAutomation amplification

   Detection

   Inbox Heartbeat Anomaly alarms fired in the US region at 19:40 UTC and declared an incident; engineers and an incident commander acknowledged it at 19:41 UTC.

   Remediation

   • PlanetScale engineers reapplied a VSchema snapshot from one hour before the incident
   • Removed the artifacts left behind by the failed MoveTables workflow
   • Rolled Vitess back to v21

   Action items

   • PlanetScale to deploy a fix for the Vitess v22 regression
   • Pause further database upgrades for Intercom until the patch is verified
   • Find the source of the packet drops in the provider's Kubernetes networking
   • Explore safeguards that protect the application from invalid topology updates
   • Review the pipeline configuration that let a branch unfreeze trigger a pending version upgrade
   • Separate maintenance tasks from version upgrades more strictly so every deployment is intended

   View case
3. IL-0369 Railway's backend API went down when a column migration waited on locks and replicas could not get database connections

   Railway2025

   Symptoms

   At 15:18 UTC on 9 December 2025, Railway's backend replicas became unhealthy and stopped serving traffic. Dashboard access, CLI operations, logins, the API and the processing of GitHub-based deployments were disrupted, while running deployments stayed online.

   Probe failuresLock contention or deadlockDatabase overloadService outage

   Trigger

   Engineers merged a database schema change with a data migration at 15:03 UTC, and the migration began applying at 15:17 UTC. It added a nullable column to a heavily used table of about 1 billion rows.

   Root cause

   Faulty deployment

   Component

   DatabaseDatabase proxy and connection poolerSchema migrationLiveness and readiness probes

   Contributing factor

   Health check tied to a dependency

   Action items

   • Increase database connections and reconfigure PgBouncer to work within those limits
   • Fix the long-running query so it no longer causes row lock issues

   View case
4. IL-0315 Railway's backend API went down when an index build locked a billion-row Postgres table and used up every connection

   Railway2025

   Symptoms

   At 18:36 UTC on 28 October 2025, two minutes after a database change went live in production, Railway's internal monitoring alerted and paged the engineering team. The dashboard could not be opened, CLI operations that use the API, such as railway up, failed, GitHub-based deployments were delayed because webhook processing failed, and all Railway Public API operations were disrupted.

   Service outageLock contention or deadlockDatabase overload

   Trigger

   A routine schema change to the backend's Postgres database, introduced at 18:23 UTC and live in production at 18:34 UTC, that added a new column with an index without the CONCURRENTLY option.

   Root cause

   Faulty deployment

   Component

   DatabaseDatabase proxy and connection pooler

   Contributing factor

   No configuration testsLimited debug access

   Detection

   Internal monitoring alerts paged the engineering team at 18:36 UTC, and the incident was declared at 18:41 UTC.

   Remediation

   • The index migration finished after about 30 minutes and released the table lock, and the queued operations and connection pools recovered

   Action items

   • Enforce CONCURRENTLY for all index creation in CI and block pull requests that do not use it
   • Adjust PgBouncer pool limits so they cannot overwhelm the Postgres instance's capacity
   • Configure database user connection limits so administrative access stays possible during incidents

   View case
5. IL-0368 PlanetScale's control plane went down and some database branches failed queries during the AWS us-east-1 outage

   PlanetScale2025

   Symptoms

   At 7:13 UTC on 20 October 2025, PlanetScale's continuous tests in production started failing because they could not create new database branches. The service that creates, resizes and configures database branches was unavailable, a near-total control plane outage, while existing database branches kept their capacity and connectivity.

   Service outageScale-up failureNetwork unreachableHigh error rateReplication lagDNS resolution failures

   Trigger

   The incident started with a DNS misconfiguration at one of PlanetScale's service providers. An Amazon DynamoDB outage impacted AWS STS, and several hours of capacity constraints and network instability in AWS us-east-1 followed.

   Root cause

   Cloud provider outage

   Component

   DatabaseDatabase replicationDatabase proxy and connection poolerObject storage

   Contributing factor

   –

   Detection

   PlanetScale's continuous testing in production started failing and alerted engineers at 7:13 UTC. At 10:05 UTC an alert fired for a Kubernetes operator in a large single-tenant customer installation exhausting its resources.

   Remediation

   • Disallowed new databases in us-east-1 and made us-east-2 the default region for new databases
   • Delayed new backups and cancelled pending backups that were waiting for an EC2 instance
   • Advised PlanetScale Managed customers with vtgate autoscaling to shed load
   • Paused draining and terminating EC2 instances older than 30 days
   • Kept vacant EC2 instances for reuse instead of terminating them
   • Bin-packed vtgate processes more tightly for customers with autoscaling
   • Sent manual reparent requests to move primaries to healthier availability zones
   • Restarted the edge load balancer and vtgate processes that did not recover after the partitions healed

   Action items

   • Understand and become resilient to the failure modes of SaaS dependencies for CI/CD, SSO, web application hosting and incident communication
   • Investigate more ambitious ways to reduce runtime dependence on internal and AWS services
   • Look into using all six us-east-1 availability zones to tolerate zonal outages and partitions

   Lessons

   Keeping the control plane separate from the data plane meant the control plane outage did not affect customer databases, and redundancy, automated failover and zonal traffic routing limited the effect of the network partitions. Network partitions are hard to reason about, test and tolerate: with three availability zones a system survives losing one zone only if the other two can still reach each other reliably.

   View case

By root cause6