Cases10 cases

1 to 5 of 10

1. IL-0572 About 10 million Pub/Sub messages backed up and Merpay accounting data was not written to Cloud Spanner for over a day

   Mercari2026

   Symptoms

   From about 6:00 on the last business day of a month, Cloud Pub/Sub messages of the Merpay accounting event pipeline began to back up, new accounting data stopped being registered in Cloud Spanner, and the accounting system intermittently could not accept new data. The backlog grew to about 10 million messages.

   Throughput degradationDatabase overloadAutoscaler misbehaviour

   Trigger

   More than a month earlier the Cloud Functions of the pipeline had been moved from 1st gen to 2nd gen (based on Cloud Run) without setting --max-instances in the deploy command, which lowered the scale limit from effectively unlimited, the 1st gen default, to 100, the 2nd gen default. The month-end peak, which on that day continued for about 40 minutes, reached that limit.

   Root cause

   Breaking change in an upgrade

   Component

   Serverless function platformMessage brokerDatabase

   Contributing factor

   Unreviewed defaultsUnknown limitChange with a delayed effectMonitoring blind spot

   Remediation

   • Re-ingested the messages saved to GCS by the Dataflow subscription, avoiding data loss

   Action items

   • Set max-instances and the other scaling parameters of Cloud Run explicitly in the deploy definition, raising max-instances from 100 to 1000 in steps
   • Set min-instances to 1 to avoid cold starts, raise concurrency from 1 to 10, and set CPU to 1 and memory from 512MB to 1Gi
   • Use the dual CPU scaling mode of spanner-autoscaler v0.8.0 with a new Total CPU threshold of 70%
   • Lower the High priority CPU threshold from 60% to 55% so scale-up starts earlier
   • Forbid Spanner scale-down between 5:00 and 7:00
   • Reserve Spanner processing units on a schedule for periods with predictable load
   • Alert in Datadog when the Cloud Run instance count reaches 40% and 60% of its limit
   • Alert in Datadog when undelivered Pub/Sub messages exceed a threshold, lowering it step by step

   Lessons

   When moving to a new generation of a managed service, defaults that the team never changed can surface as problems under load, so parameters that only matter at peak, such as scale limits, concurrency and timeouts, should be set explicitly at every migration. An outage rarely has one cause; the fact that the previous month passed under the same conditions pointed to traffic continuity, and prevention needs to look at the trigger, the amplification and what turned it into damage. Capacity settings and the monitoring that shows when they are approached have to be designed together, and teams that run systems they inherited should review these settings and assumptions regularly.

   View case
2. IL-0435 Error rates rose as Aurora MySQL killed connections and declined queries to avoid OOM while load metrics stayed flat

   CyberAgent2026

   Symptoms

   An alert fired one day because an application's error rate had suddenly risen. The error log of the Aurora MySQL RDS instance for that time showed Aurora killing connections and declining queries to avoid running out of memory. The killed queries also made application latency worse.

   High error rateMemory pressure

   Trigger

   An application release added a batch job run every minute by a Kubernetes CronJob. Each run's short-lived pod connected to the database from a new pod IP.

   Root cause

   Capacity exhaustion

   Component

   DatabaseJobs and CronJobs

   Contributing factor

   Unreviewed defaultsChange with a delayed effect

   Detection

   An alert fired on the application's error rate.

   Remediation

   • Set performance_schema to manual management and capped two related Performance Schema parameters, unlimited by default, at 100 in the parameter group, then restarted the RDS instance

   Lessons

   When system alerts fire but the database's load metrics look calm, Performance Schema growth is worth suspecting. The two parameters that were capped are unlimited by default, so connections from ever-changing source addresses keep adding to the records. Capping them has a cost: once the cap is exceeded, per-host figures in Performance Insights are missing. Putting RDS Proxy between the application and Aurora MySQL to keep source hosts stable is another possible measure.

   View case
3. IL-0288 ZooKeeperWait rose to thousands of seconds on three redundant ClickHouse instances, degrading a CDN77 data pipeline

   CDN772026

   Symptoms

   At 13:30 UTC on 9 March 2026, CDN77's data team was alerted to a sudden rise in the ZooKeeperWait metric on the ClickHouse cluster behind one of its large data pipelines. Alerting for this cluster size is set at 10 sec for the low watermark and 20 sec for the high one, and the metric reached thousands of seconds. The same rise appeared on three symmetrically redundant ClickHouse instances in different datacenters, each with its own ZooKeeper instance, and the pipeline stayed degraded for over an hour, with no observable impact on internal users, customers or partners.

   Replication lagThroughput degradation

   Trigger

   From 13:15 UTC two client machines that had been cut off from the network for several months came back and their log shippers sent months of queued, very small batches of health check data with old timestamps, spanning a very large time range.

   Root cause

   Capacity exhaustion

   Component

   DatabaseCoordination and key-value storeDatabase replicationTelemetry agent

   Contributing factor

   Monitoring blind spotUnreviewed defaultsIgnored early signal

   Detection

   An alert on the ZooKeeperWait metric fired at 13:30 UTC, fifteen minutes after the problematic inserts began at 13:15. The cluster's alerting uses a low watermark of 10 sec and a high watermark of 20 sec.

   Remediation

   • Dropped and recreated the one table partitioned by 5 minutes that had not cleared its backlog of parts
   • Broadened that table's partitioning from minutes to hours

   Action items

   • Reject out-of-range timestamps in the root MaterializedView tables
   • Record statistics on rejected rows in a second materialized view
   • Refine the deliberately wide rejection boundaries as data on rejections comes in
   • Keep a minimal test scenario with an insert query whose timestamps span days or months
   • Make a range of observability improvements

   Lessons

   Data structure matters even in a pipeline built around volume: incoherent timestamps can destabilise it as effectively as changes in cardinality. A metric that looks odd and cannot be explained at once should not be set aside under incident pressure; the falling mean insert size was what led to the cause. Means and medians each have their place, and here the mean, not the median, exposed the problem. Partitioning has to be chosen carefully: too few partitions give oversized parts, slow mutations and costly or ineffective TTL, while too many slow column scans, add final merging to aggregating queries and overload the Keeper.

   View case
4. IL-0316 Railway's control plane slowed and deployments stopped after a PgBouncer version jump used abnormal CPU and memory

   Railway2025

   Symptoms

   At 11:48 UTC on 22 September 2025, minutes after Railway had resolved a dashboard outage, its control plane showed abnormal resource consumption and elevated API latency. A large backlog of deployments queued during the first outage had started all at once, and the control plane began returning elevated database errors.

   High latencyIntermittent timeoutsDeploys failing

   Trigger

   After a dashboard outage caused by a schema change was fixed, a large backlog of queued deployments started at the same time and put the control plane and its database, behind a newly upgraded PgBouncer, under heavy load.

   Root cause

   Breaking change in an upgrade

   Component

   DatabaseDatabase proxy and connection pooler

   Contributing factor

   Unpinned versionNo staged rolloutMonitoring blind spotUnreviewed defaults

   Detection

   Users reported at 11:05 UTC that the project canvas failed to load; the control plane's abnormal resource consumption and elevated API latency were seen at 11:48 UTC and opened as a new incident.

   Remediation

   • Deployed a fix for the API failure caused by the schema change within 15 minutes
   • Suspended deployments for non-Pro users so Pro users could deploy and the queue could drain
   • Suspended all deployments, Pro included, for up to 20 minutes during peak instability
   • Scaled up the control plane database's CPU, memory, connection limits and IOPS
   • Rolled PgBouncer back to the previous stable version, which brought all metrics back to nominal
   • Re-enabled deployments for all non-Pro users
   • Fixed the edge case in which private networks and their endpoints were missing from the dashboard

   Action items

   • Re-engineer parts of the control plane to be more resilient, moving large parts of its storage layer away from Postgres
   • Add CI validation that keeps disruptive schema changes out of production and automated schema compatibility checks before backend deployments
   • Monitor and alert on PgBouncer node CPU, memory and connection pool utilization
   • Upgrade PgBouncer in phases, instance by instance, instead of everywhere at once

   Lessons

   Railway believes that without the breaking schema change reaching production the cascade would not have happened, that monitoring PgBouncer's own CPU, memory and pool metrics would have shown the performance regression earlier and shortened recovery, and that upgrading PgBouncer instances one at a time would have exposed the version-specific problem before it reached every instance.

   View case
5. IL-0436 A supremum lock left by ROLLBACK TO SAVEPOINT in a long transaction blocked inserts and filled Aurora MySQL connections

   freee2025

   Symptoms

   Connections to Aurora MySQL failed again and again in one freee service until its web pages could not be opened at all. The database did not respond, the web application timed out and could not open new database connections, and the whole service was down.

   Lock contention or deadlockService outage

   Trigger

   A long transaction inserted a row into the table after a SAVEPOINT and then ran ROLLBACK TO SAVEPOINT, which left it holding an exclusive lock on the supremum of the table's primary index.

   Root cause

   Software defect

   Component

   Database

   Contributing factor

   Unverified assumptionUnreviewed defaults

   Action items

   • As a rule stop issuing SAVEPOINT, split a transaction when a partial rollback is needed, and set an application-side timeout where a nested transaction is unavoidable
   • Teach developers to consider READ COMMITTED per session for use cases that can accept phantom rows
   • Monitor long transactions and keep working to remove them across the company

   Lessons

   ROLLBACK TO SAVEPOINT rolls back rows but does not release their locks, so rolling back one inserted row can leave a supremum lock that blocks every INSERT into the table. ORM frameworks, including Active Record for nested transactions, use SAVEPOINT internally without developers noticing. Locks held by long transactions often go unnoticed when few workloads are blocked, and this incident showed their hidden risk.

   View case

By root cause5