Kafka Error SNAPSHOT_NOT_FOUND
Error code 98 · Non-retriable Admin
Requested snapshot was not found.
Common Causes
- A KRaft follower or observer requested a snapshot by offset/epoch that no longer exists on the leader — the snapshot was deleted as part of log retention
- New broker joining the cluster requested a snapshot for bootstrapping, but the leader has not yet generated a snapshot at the requested point
- FetchSnapshot request was issued with an incorrect snapshot ID (wrong end offset or epoch) due to a stale reference
Solutions
- Allow the KRaft leader to generate a new snapshot — it does so periodically based on metadata.log.max.record.bytes.between.snapshots; ensure this is not set too high
- If a broker is stuck in snapshot fetch, restart it to trigger a fresh snapshot request with the latest available snapshot ID from the leader
- Check the leader's snapshot directory (__cluster_metadata-0/) to confirm available snapshots and their offsets
Diagnostic Commands
# List KRaft checkpoint files on disk
ls -lh /var/kafka/data/__cluster_metadata-0/*.checkpoint 2>/dev/null || ls -lh /var/kafka/data/__cluster_metadata-0/
# Check KRaft metadata quorum status
kafka-metadata-quorum.sh --bootstrap-server localhost:9092 describe --status | grep -E 'HighWatermark|LogEndOffset|Snapshot'Related APIs
This error can be returned by: FetchSnapshot
Debugging Kafka errors? Conduktor Console gives you real-time visibility into your cluster. Explore all errors in the Error Decoder.