Aiven incident

Kafka upgrade to 3.9 paused

Minor Resolved View vendor source →

Aiven experienced a minor incident on June 25, 2026, lasting —. The incident has been resolved; the full update timeline is below.

Started
Jun 25, 2026, 11:38 AM UTC
Resolved
Jun 25, 2026, 11:38 AM UTC
Duration
Detected by Pingoru
Jun 25, 2026, 11:38 AM UTC

Update timeline

  1. investigating Jun 23, 2026, 02:18 PM UTC

    Status: Investigating We have paused Kafka 3.8 ->3.9 upgrades as we are investigating an issue with KRaft migration during the upgrade. Aiven is actively monitoring affected services and applying mitigations where required. Impact: Existing Kafka services continue to run. Customers attempting to upgrade Kafka services from version 3.8 to 3.9 will not be able to complete the upgrade while KRaft migrations are paused. Next update: We will provide an update once more information is available.

  2. identified Jun 23, 2026, 03:10 PM UTC

    Status: Identified We have identified two separate problems: • An issue in migration code • An upstream bug: https://github.com/apache/kafka/pull/19854 We are working on the fix for the migration issue and back-porting the fix for the upstream bug.

  3. identified Jun 24, 2026, 03:01 PM UTC

    Status: Identified We have identified issues affecting KRaft migrations (Kafka 3.8 to 3.9 upgrades). We have paused all migrations while we work on a fix. If you are experiencing any issues, please contact support.

  4. monitoring Jun 25, 2026, 08:06 AM UTC

    Status: Monitoring We have deployed fixes for the issues affecting KRaft migrations during Kafka 3.8 to 3.9 upgrades, and migrations have been re-enabled. We are continuing to monitor the situation to ensure services remain stable. If you are experiencing any issues, please contact support.

  5. resolved Jun 25, 2026, 11:38 AM UTC

    Status: Resolved The issues affecting KRaft migrations during Kafka 3.8 to 3.9 upgrades have been resolved. Fixes have been deployed, migrations have been re-enabled, and services have remained stable during monitoring. If you are experiencing any issues, please contact support.