← Back to the carnage
COCKROACHDB
🔴 Two availability zones decided to ghost me-central-1 entirely while me-south-1 is limping along on life support. AWS infrastructure apparently scheduled unplanned maintenance with all the grace of a cockroach under a shoe. (Auto-resolved: no updates in 24h)
// The official story
Cluster unavailability and degraded cluster operations in AWS me-central-1 and me-south-1 regions
// Status page says
Mar 2 , 09:01 UTC Update - me-central-1 region is impacted more with two AZs out and the clusters in that region are unavailable. In addition to this, me-south-1 region is also experiencing an AZ outage, which causes degraded performance of clusters in me-south-1. We are closely tracking AWS's status. Impacted customers can track AWS status here https://health.aws.amazon.com/health/status Mar 2 , 02:47</v
🔗 Official status page (for the PR-approved version)
https://status.cockroachlabs.cloud/incidents/fhx8k1nh1kz7Cockroachdb's Greatest Hits
🟢 COCKROACHDB
Sep 01, 2026
🟠 Google Cloud's infrastructure threw a tantrum, leaving Cockroach Cloud customers in us-central-1 and asia-east-1 dealing with latency, delays, and connection issues. Turns out blaming 'the cloud' is literally accurate this time. Teams are 'fully engaged' which is corporate for 'sweating profusely.'. (Auto-resolved: no updates in 24h)
🟢 COCKROACHDB
Mar 18, 2026
🔴 COCKROACHDB - Cloud Console Social Login Auth Unavailable. (Auto-resolved: no updates in 24h)
🟢 COCKROACHDB
Mar 11, 2026
⚪ AWS me-central-1 had a very bad day when someone tripped over a power cable in the UAE data center, yeeting multiple availability zones offline. CockroachDB clusters in the region went full zombie mode while me-south-1 just limped along like it'd been hit by a truck. AWS is still picking up the pieces. (Auto-resolved: no updates in 24h)
🟢 COCKROACHDB
Mar 02, 2026
🔴 Two availability zones collapsed in AWS me-central-1, leaving clusters completely offline. me-south-1 got the degraded experience special with single AZ failure. GCP and Azure laughing from the sidelines while customers frantically dust off disaster recovery plans written in 2019. (Auto-resolved: no updates in 24h)