nodetool listsnapshots¶

Displays a list of all snapshots on the node with their sizes.

Synopsis¶

nodetool [connection_options] listsnapshots [-e] [-n]

Description¶

nodetool listsnapshots shows all snapshots on the local node, including:

Snapshot tag name
Keyspace and table names
True data size
Disk space consumed

Essential for monitoring backup state and managing disk space.

Options¶

Option	Description
`-e, --ephemeral`	Include ephemeral snapshots
`-n, --no-ttl`	Exclude snapshots with TTL

Output Example¶

Snapshot Details:
Snapshot name       Keyspace name      Column family name   True size   Size on disk   Creation time
backup_20240115     my_keyspace        users                1.5 GB      1.5 GB         2024-01-15T10:30:00Z
backup_20240115     my_keyspace        orders               2.3 GB      2.3 GB         2024-01-15T10:30:00Z
backup_20240108     my_keyspace        users                1.2 GB      0 bytes        2024-01-08T10:30:00Z
backup_20240108     my_keyspace        orders               2.1 GB      0 bytes        2024-01-08T10:30:00Z
pre_upgrade         system_schema      tables               45 MB       45 MB          2024-01-14T22:00:00Z

Total TrueDiskSpaceUsed: 7.15 GB

Output Fields¶

Field	Description
Snapshot name	Tag used when creating the snapshot
Keyspace name	Keyspace containing the table
Column family name	Table name
True size	Actual data size in the snapshot
Size on disk	Current disk space consumed
Creation time	When the snapshot was created

Understanding Size Fields¶

True size	Size on disk	Meaning
1.5 GB	1.5 GB	Original SSTables still exist (hard links)
1.5 GB	0 bytes	Original SSTables compacted away (space freed)
1.5 GB	500 MB	Partially compacted

Size on Disk

"Size on disk" shows actual disk blocks consumed. When original SSTables are compacted, the hard links become the only reference, and size on disk increases.

Examples¶

List All Snapshots¶

nodetool listsnapshots

Include Ephemeral Snapshots¶

nodetool listsnapshots -e

Exclude TTL Snapshots¶

nodetool listsnapshots -n

Filter by Snapshot Name¶

nodetool listsnapshots | grep backup_20240115

Total Snapshot Size¶

nodetool listsnapshots | tail -1

Count Snapshots¶

nodetool listsnapshots | grep -c "^[a-zA-Z]"

When to Use¶

Before Creating New Snapshots¶

Check existing snapshots to avoid duplicates:

nodetool listsnapshots | grep my_new_tag

Disk Space Investigation¶

When disk usage is high:

nodetool listsnapshots
# Check Total TrueDiskSpaceUsed at the bottom

Backup Verification¶

Verify snapshot was created:

nodetool snapshot -t my_backup my_keyspace
nodetool listsnapshots | grep my_backup

Cleanup Planning¶

Identify snapshots to remove:

nodetool listsnapshots
# Find old snapshots by creation time

Interpreting Results¶

Recently Created Snapshot¶

backup_20240115     my_keyspace     users     1.5 GB     1.5 GB     2024-01-15T10:30:00Z

True size = Size on disk = Original SSTables still exist
Hard links share disk blocks with active SSTables
Minimal additional space used

Aged Snapshot¶

backup_20240101     my_keyspace     users     1.5 GB     1.5 GB     2024-01-01T10:30:00Z

After compaction runs:

Original SSTables removed
Snapshot files become sole reference
Now consuming full 1.5 GB

No Snapshots¶

Snapshot Details:

Total TrueDiskSpaceUsed: 0 bytes

No snapshots exist on this node.

Monitoring Snapshot Growth¶

Check All Nodes¶

for node in node1 node2 node3; do
    echo "=== $node ==="
    ssh "$node" "nodetool listsnapshots | tail -3"
done

Track Over Time¶

# Log snapshot sizes daily
echo "$(date),$(nodetool listsnapshots | tail -1 | awk '{print $2}')" >> /var/log/snapshot_sizes.csv

Alert on Large Snapshots¶

total=$(nodetool listsnapshots | tail -1 | awk '{print $2}' | sed 's/GB//')
if (( $(echo "$total > 100" | bc -l) )); then
    echo "WARNING: Snapshot usage exceeds 100GB"
fi

Snapshot Location on Disk¶

Snapshots are stored in:

/var/lib/cassandra/data/<keyspace>/<table>-<uuid>/snapshots/<tag>/

To manually check sizes:

# All snapshots
du -sh /var/lib/cassandra/data/*/*/snapshots/*

# Specific snapshot
du -sh /var/lib/cassandra/data/*/*/snapshots/backup_20240115/

Common Issues¶

Empty Output¶

If no snapshots are shown but you expect them:

Check if snapshots were taken on this node

Verify snapshot directory exists:

ls /var/lib/cassandra/data/*/*/snapshots/

Incorrect Size Reported¶

Size reporting depends on filesystem support for hard links:

Some filesystems may not accurately report shared blocks
Use du for precise measurements if needed

Very Large Total Size¶

Old snapshots consuming significant space:

# Identify largest snapshots
nodetool listsnapshots | sort -k4 -h

# Clear old snapshots
nodetool clearsnapshot -t old_snapshot_tag

Best Practices¶

Snapshot Monitoring Guidelines

Regular checks - Include in daily monitoring
Set retention alerts - Alert when snapshots age
Track total size - Monitor growth trends
Clean up promptly - Remove snapshots after backup verification
Use TTL snapshots - Auto-delete when possible (4.0+)
Document naming - Use consistent, dated snapshot names

Command	Relationship
snapshot	Create snapshots
clearsnapshot	Remove snapshots
tablestats	Table-level statistics