nodetool snapshot

Creates a backup by taking a snapshot of table data. A snapshot is a hardlink to the SSTable files in the data directory for a schema table at the moment the snapshot is executed.

The snapshot directory path is: data/<keyspace_name>/<table-UUID>/snapshots/<snapshot_name>. Data is backed up into multiple .db files and table schema is saved to schema.cql. The schema.cql file captures the structure of the table at the time of snapshot. Restoring the snapshot requires the table to have the same structure. See Manual Backup and Restore, with Point-in-time and table-level restore.

Always run nodetool cleanup before taking a snapshot for restore. Otherwise invalid replicas, that is replicas that have been superseded by new, valid replicas on newly added nodes can get copied to the target when they should not. This results in old data showing up on the target.

Before upgrading Hyper-Converged Database (HCD), be sure to create a backup of all keyspaces. See Taking a snapshot.

Synopsis

nodetool [<connection_options>] snapshot
[--table <table_name> | -kt keyspace_name.table_name,...]
[-sf] [-t <snapshotname>] [--]
[<keyspace_name> [<keyspace_name>...]]

Definition

The short- and long-form options are comma-separated.

Connection options

-h, --host hostname: The hostname or IP address of a remote node or nodes. When omitted, the default is the local machine.
-p, --port jmx_port: The JMX port number.
-pw, --password jmxpassword: The JMX password for authenticating with secure JMX. If a password is not provided, you are prompted to enter one.
-pwf, --password-file jmx_password_filepath: The file path to the file that stores JMX authentication credentials.
-u, --username jmx_username: The username for authenticating with secure JMX.

Command arguments

--

Separates an option from an argument that could be mistaken for an option.

--table, -cf, --column-family table_name

Table name in the specified keyspace.

-kt, --kt-list, -kc, --kc.list keyspace_name.table_name,…

Comma-separated list of keyspace_name.table_name with no spaces after the comma. For example: cycling.cyclist,basketball.players

-sf, --skip_flush

Do not flush tables before creating the snapshot.

Snapshot will not contain unflushed data.

-t snapshotname, --tag snapshotname

The snapshot file path. When not specified, the current time is used for the directory name. For example, 1489076973698.

Examples

Take a snapshot of all keyspaces on the node

Run nodetool cleanup before taking the snapshot:

nodetool cleanup

Take a snapshot of all keyspaces:

nodetool snapshot

+ Results include the name of the snapshot directory:

Requested creating snapshot(s) for [all keyspaces] with snapshot name [1489076973698] and options {skipFlush=false}
Snapshot directory: 1489076973698

Create a tagged snapshot of keyspace

Before taking the snapshot, run nodetool cleanup.

Take tagged snapshot of cycling keyspace in the cycling_2017-3-9 directory:

nodetool snapshot -t cycling_2017-3-9 cycling

Results:

Requested creating snapshot(s) for [cycling] with snapshot name [2015.07.17] and options {skipFlush=false}
Snapshot directory: cycling_2017-3-9

The cycling keyspace contains two tables, cyclist_name and upcoming_calendar. The snapshot creates multiple snapshot directories named cycling_2017-3-9. A number of .db files containing the data are located in these directories, along with table schema. For example, from the HCD installation directory:

ls -1 data/cycling/cyclist_name-9e516080f30811e689e40725f37c761d/snapshots/cycling_2017-3-9

manifest.json
mc-1-big-CompressionInfo.db
mc-1-big-Data.db
mc-1-big-Digest.crc32
mc-1-big-Filter.db
mc-1-big-Index.db
mc-1-big-Statistics.db
mc-1-big-Summary.db
mc-1-big-TOC.txt
schema.cql

Take a snapshot of multiple keyspaces

Before taking the snapshot, run nodetool cleanup.

Take snapshot of the mykeyspace and cycling keyspaces:

nodetool snapshot mykeyspace cycling

Results:

Requested creating snapshot(s) for [mykeyspace, cycling] with snapshot name [1391460334889] and options {skipFlush=false}
Snapshot directory: 1391460334889

Take a snapshot of a single table

Before taking the snapshot, run nodetool cleanup.

Take a snapshot of the cyclist_name table in the cycling keyspace:

nodetool snapshot --table cyclist_name cycling

Results:

Requested creating snapshot(s) for [cycling] with snapshot name [1391461910600] and options {skipFlush=false}
Snapshot directory: 1391461910600

The resulting snapshot directory 1391461910600 contains data files and the schema of cyclist_name table in data/cycling/cyclist_name-a882dca02aaf11e58c7b8b496c707234/snapshots.

Take a snapshot of multiple tables in different keyspaces

Before taking the snapshot, run nodetool cleanup.

Take a snapshot the cyclist_name table in the cycling keyspace and the sample_times table in the test keyspace:

nodetool snapshot -kt cycling.cyclist_name,test.sample_times

Results:

Requested creating snapshot(s) for [cycling.cyclist_name,test.sample_times] with snapshot name [1431045288401] and options {skipFlush=false}
Snapshot directory: 1431045288401

Create a snapshot of keyspace without flushing tables

Before taking the snapshot, run nodetool cleanup.

Take snapshot of cycling keyspace with -sf option:

nodetool snapshot cycling -sf

Results:

Requested creating snapshot(s) for [cycling] with snapshot name [1431045288401] and options {skipFlush=true}
Snapshot directory: 1431045288401

nodetool snapshot

Synopsis

Definition

Connection options

Command arguments

Examples

Was this helpful?

Give Feedback