Questions
Status
Queue depth increases tail latencyHardOn-call incidentGoogle styleOn-call incidentGoogle styleHard35 minConsumer lag on few partitionsHardOn-call incidentOn-call incidentHard35 minCache serialization format collisionHardOn-call incidentOn-call incidentHard40 minConsumer stalls on schema mismatchHardOn-call incidentOn-call incidentHard40 minVendor schema change silent failureHardOn-call incidentStripe styleOn-call incidentStripe styleHard40 minPartial network partition splits quorumHardOn-call incidentOn-call incidentHard45 minFailover controller creates split-brainHardOn-call incidentOn-call incidentHard45 minOCSP responder timeout blocks handshakesHardOn-call incidentOn-call incidentHard40 minFeature store endianness mismatchHardOn-call incidentOn-call incidentHard40 minLRU evicts durable keysHardOn-call incidentStripe styleOn-call incidentStripe styleHard35 minTTL backfill deletes active documentsHardOn-call incidentAmazon styleOn-call incidentAmazon styleHard35 minMigrated accounts show stale balancesHardOn-call incidentStripe styleOn-call incidentStripe styleHard40 minCache invalidation storm from backfillHardOn-call incidentOn-call incidentHard35 minConsumer stuck on oversized recordHardOn-call incidentOn-call incidentHard35 minSessions expire ahead of scheduleHardOn-call incidentOn-call incidentHard35 minCross-tenant data leakHardOn-call incidentSalesforce styleOn-call incidentSalesforce styleHard40 minCascade delete kills child recordsHardOn-call incidentOn-call incidentHard35 minCredential stuffing evades login alarmsHardOn-call incidentPayPal styleOn-call incidentPayPal styleHard40 minService account reads spike sequentiallyHardOn-call incidentOn-call incidentHard40 minCache busting attackHardOn-call incidentOn-call incidentHard40 minBackup repository encryption detectedHardOn-call incidentOn-call incidentHard45 minRansomware staging before detonationHardOn-call incidentOn-call incidentHard45 minSecret rotation breaks webhook verificationHardOn-call incidentStripe styleOn-call incidentStripe styleHard40 minOver-broad OAuth scopes issuedHardOn-call incidentOn-call incidentHard40 minService account off-hours vault accessHardOn-call incidentOn-call incidentHard40 minRedirect bypass to instance metadataHardOn-call incidentAmazon styleOn-call incidentAmazon styleHard45 minSession fixation after loginHardOn-call incidentOn-call incidentHard40 minDead-letter topic overflowHardOn-call incidentStripe styleOn-call incidentStripe styleHard40 minRebalance stops offset commitsHardOn-call incidentOn-call incidentHard40 minUneven partition assignment after scaleHardOn-call incidentOn-call incidentHard40 minTopic retention deletes unconsumed messagesHardOn-call incidentOn-call incidentHard40 minExport all records OOMKilledHardOn-call incidentOn-call incidentHard30 minIndex merge swap thrashingHardOn-call incidentOn-call incidentHard35 minConfig reload file descriptor leakHardOn-call incidentOn-call incidentHard35 minHot partition index bloatHardOn-call incidentOn-call incidentHard35 minJIT deoptimization cliffHardOn-call incidentOn-call incidentHard40 minControl plane requests blockedHardOn-call incidentOn-call incidentHard40 minHTTP/2 stream head-of-line blockingHardOn-call incidentGoogle styleOn-call incidentGoogle styleHard40 minHedged request amplificationHardOn-call incidentGoogle styleOn-call incidentGoogle styleHard40 minDirty page writeback stallHardOn-call incidentOn-call incidentHard40 minSafepoint spin time latencyHardOn-call incidentOn-call incidentHard40 minNetwork bandwidth burst exhaustionHardOn-call incidentAmazon styleOn-call incidentAmazon styleHard40 minTLS session cache miss spikeHardOn-call incidentOn-call incidentHard40 minRetries hammer degraded nodeHardOn-call incidentOn-call incidentHard40 minFlag evaluation percentage drifts upwardHardOn-call incidentDoorDash styleOn-call incidentDoorDash styleHard40 minCanary cache key cardinality explosionHardOn-call incidentOn-call incidentHard40 minIdempotency key version skewHardOn-call incidentStripe styleOn-call incidentStripe styleHard40 minMemory leak undetected in short canaryHardOn-call incidentOn-call incidentHard40 minOmitted field null deserializationHardOn-call incidentOn-call incidentHard40 minLong-running matches evicted during rolloutHardOn-call incidentOn-call incidentHard40 min