Cluster commands

This is archived documentation for InfluxData product versions that are no longer maintained. For newer documentation, see the latest InfluxData documentation.

Use the command line tools influxd-ctl and influx to interact with your InfluxDB Enterprise cluster and data.

Content

influxd-ctl

The influxd-ctl tool is available on all meta nodes. Use influxd-ctl to manage your cluster nodes, backup data, restore data, and rebalance your cluster.

Syntax

influxd-ctl [global-options] <command> [arguments]

Global Options

-auth-type [ none | basic | jwt ]    Specify the type of authentication to use. The default authentication type is none.

-bind <hostname>:<port>    Specify the bind HTTP address of a meta node to connect to. The default is localhost:8091.

-bind-tls    Use TLS. If you have enabled HTTPS, you MUST use this option in order for influxd-ctl to connect to the meta node.

-config '<path-to-configuration-file>'    Specify the path to the configuration file.

-pwd <password>    Specify the user’s password. This option is ignored if -auth-type basic isn’t specified.

-k    Skip certificate verification; use this option with a self-signed certificate. -k is ignored if -bind-tls isn’t specified.

-secret <JWT-shared-secret>    Specify the JSON Web Token (JWT) shared secret. This option is ignored if -auth-type jwt isn’t specified.

-user <username>    Specify the user’s username. This option is ignored if -auth-type basic isn’t specified.

Examples of Global Options

The following examples use the tool’s show argument. See the section below for more information about influxd-ctl’s arguments.

Example 1: Bind to a remote meta node

$ influxd-ctl -bind meta-node-02:8091 show

The tool binds to the meta node with the hostname meta-node-02 at port 8091. By default, the tool binds to the meta node with the hostname localhost at port 8091.

Example 2: Authenticate with JWT

$ influxd-ctl -auth-type jwt -secret oatclusters show

The tool uses JWT authentication with the shared secret oatclusters.

If authentication is enabled in the cluster’s meta node configuration files and data node configuration files and the influxd-ctl command does not include authentication details, the system returns:

Error: unable to parse authentication credentials.

If authentication is enabled and the influxd-ctl command provides the incorrect shared secret, the system returns:

Error: signature is invalid.

Example 3: Authenticate with basic authentication

$ influxd-ctl -auth-type basic -user admini -pwd mouse show

The tool uses basic authentication for the cluster user admini with the password mouse.

If authentication is enabled in the cluster’s meta node configuration files and data node configuration files and the influxd-ctl command does not include authentication details, the system returns:

Error: unable to parse authentication credentials.

If authentication is enabled and the influxd-ctl command provides the incorrect username or password, the system returns:

Error: authorization failed.

Arguments

add-data

Adds a data node to a cluster. By default, influxd-ctl adds the specified data node to the local meta node’s cluster. Use add-data instead of the join argument when performing a Production Installation of an InfluxDB Enterprise cluster.

add-data <data-node-TCP-bind-address>

Resources: Production Installation

Examples


Example 1: Add a data node to a cluster using the local meta node

$ influxd-ctl add-data cluster-data-node:8088

Added data node 3 at cluster-data-node:8088

The command contacts the local meta node running at localhost:8091 and adds a data node to that meta node’s cluster. The data node has the hostname cluster-data-node and runs on port 8088.

Example 2: Add a data node to a cluster using a remote meta node

$ influxd-ctl -bind cluster-meta-node-01:8091 add-data cluster-data-node:8088

Added data node 3 at cluster-data-node:8088

The command contacts the meta node running at cluster-meta-node-01:8091 and adds a data node to that meta node’s cluster. The data node has the hostname cluster-data-node and runs on port 8088.

add-meta

Adds a meta node to a cluster. By default, influxd-ctl adds the specified meta node to the local meta node’s cluster. Use add-meta instead of the join argument when performing a Production Installation of an InfluxDB Enterprise cluster.

add-meta <meta-node-HTTP-bind-address>

Resources: Production Installation

Examples


Example 1: Add a meta node to a cluster using the local meta node

$ influxd-ctl add-meta cluster-meta-node-03:8091

Added meta node 3 at cluster-meta-node:8091

The command contacts the local meta node running at localhost:8091 and adds a meta node to that local meta node’s cluster. The added meta node has the hostname cluster-meta-node-03 and runs on port 8091.

Example 2: Add a meta node to a cluster using a remote meta node

$ influxd-ctl -bind cluster-meta-node-01:8091 add-meta cluster-meta-node-03:8091

Added meta node 3 at cluster-meta-node-03:8091

The command contacts the meta node running at cluster-meta-node-01:8091 and adds a meta node to that meta node’s cluster. The added meta node has the hostname cluster-meta-node-03 and runs on port 8091.

backup

Creates a backup of a cluster’s metastore and shard data at that point in time and stores the copy in the specified directory. Backups are incremental by default; they create a copy of the metastore and shard data that have changed since the previous incremental backup. If there are no existing incremental backups, the system automatically performs a complete backup.

backup [ -db <database> | -from <data-node-TCP-bind-address> | -full | -rp <retention-policy> | -shard <shard-id> ] <backup-directory>

Options:

-db <database>:    The name of the single database to back up.

-from <data-node-TCP-address>:    The TCP address of the target data node.

-full:    Perform a full backup.

-rp <retention-policy>:    The name of the single retention policy to back up (requires the -db flag).

-shard <shard-id>:    The ID of the single shard to back up.

Restoring a -full backup and restoring an incremental backup require different syntax. To prevent issues with restore, keep -full backups and incremental backups in separate directories.

Resources: Backup and Restore

Examples


Example 1: Perform an incremental backup

$ influxd-ctl backup .

Output:

Backing up meta data... Done. 421 bytes transferred
Backing up node cluster-data-node:8088, db telegraf, rp autogen, shard 4... Done. Backed up in 903.539567ms, 307712 bytes transferred
Backing up node cluster-data-node:8088, db _internal, rp monitor, shard 1... Done. Backed up in 138.694402ms, 53760 bytes transferred
Backing up node cluster-data-node:8088, db _internal, rp monitor, shard 2... Done. Backed up in 101.791148ms, 40448 bytes transferred
Backing up node cluster-data-node:8088, db _internal, rp monitor, shard 3... Done. Backed up in 144.477159ms, 39424 bytes transferred
Backed up to . in 1.293710883s, transferred 441765 bytes

$ ls
20160803T222310Z.manifest  20160803T222310Z.s1.tar.gz  20160803T222310Z.s3.tar.gz
20160803T222310Z.meta      20160803T222310Z.s2.tar.gz  20160803T222310Z.s4.tar.gz

The command performs an incremental backup and stores it in the current directory. If there are any existing backups the current directory, the system performs an incremental backup. If there aren’t any existing backups in the current directory, the system performs a complete backup of the cluster.

Example 2: Perform a full backup

$ influxd-ctl backup -full backup_dir

Output:

Backing up meta data... Done. 481 bytes transferred
Backing up node cluster-data-node:8088, db _internal, rp monitor, shard 1... Done. Backed up in 33.207375ms, 238080 bytes transferred
Backing up node cluster-data-node:8088, db telegraf, rp autogen, shard 2... Done. Backed up in 15.184391ms, 95232 bytes transferred
Backed up to backup_dir in 51.388233ms, transferred 333793 bytes

~# ls backup_dir
20170130T184058Z.manifest
20170130T184058Z.meta
20170130T184058Z.s1.tar.gz
20170130T184058Z.s2.tar.gz

The command performs a full backup of the cluster and stores the backup in the existing directory backup_dir.

copy-shard

Copies a shard from a source data node to a destination data node.

copy-shard <data-node-source-TCP-address> <data-node-destination-TCP-address> <shard-id>

Resources: Cluster Rebalance

Examples


Example 1: Copy a shard from one data node to another data node

$ influxd-ctl copy-shard cluster-data-node-01:8088 cluster-data-node-02:8088 22'

Copied shard 22 from cluster-data-node-01:8088 to cluster-data-node-02:8088

The command copies the shard with the id 22 from the data node running at cluster-data-node-01:8088 to the data node running at cluster-data-node-02:8088.

copy-shard-status

Shows all in-progress copy shard operations, including the shard’s source node, destination node, database, retention policy, shard id, total size, current size, and the operation’s start time.

copy-shard-status
Examples


Example 1: Show all in-progress copy-shard operations

$ influxd-ctl copy-shard-status

Source                     Dest                       Database  Policy   ShardID  TotalSize  CurrentSize  StartedAt
cluster-data-node-02:8088  cluster-data-node-03:8088  telegraf  autogen  34       119624324  119624324    2017-06-22 23:45:09.470696179 +0000 UTC

The command returns one in-progress copy-shard operation. The system is copying shard 34 from cluster-data-node-02:8088 to cluster-data-node-03:8088. Shard 34 is associated with the telegraf database and the autogen retention policy. The TotalSize and CurrentSize columns are reported in bytes.

join

Joins a meta node and/or data node to a cluster. By default, influxd-ctl joins the local meta node and/or data node into a new cluster. Use join instead of the add-meta or add-data arguments when performing a QuickStart Installation of an InfluxDB Enterprise cluster.

join [-v] [<meta-node-HTTP-bind-address>]

Options:

-v    Prints verbose information about the join.

meta-node-HTTP-bind-address    The address of a meta node in an existing cluster. Use this option to add the un-joined meta node and/or data node to an existing cluster.

Resources: QuickStart Installation

Examples


Example 1: Join a meta and data node into a cluster

$ influxd-ctl join

Joining meta node at localhost:8091
Searching for meta node on cluster-node-03:8091...
Searching for data node on cluster-node-03:8088...

Successfully created cluster

  * Added meta node 1 at cluster-node-03:8091
  * Added data node 2 at cluster-node-03:8088

  To join additional nodes to this cluster, run the following command:

  influxd-ctl join cluster-node-03:8091

The command joins the meta node running at cluster-node-03:8091 and the data node running at cluster-node-03:8088 into a new cluster.

Example 2: Join a meta and data node to an existing cluster

$ influxd-ctl join cluster-meta-node-02:8091

Joining meta node at cluster-meta-node-02:8091
Searching for meta node on cluster-node-03:8091...
Searching for data node on cluster-node-03:8088...

Successfully joined cluster

  * Added meta node 3 at cluster-node-03:8091
  * Added data node 4 at cluster-node-03:8088

The command joins the meta node running at cluster-node-03:8091 and the data node running at cluster-node-03:8088 to an existing cluster. The existing cluster includes the meta node running at cluster-meta-node-02:8091.

Example 3: Join a meta node to an existing cluster

$ influxd-ctl join cluster-meta-node-02:8091

Joining meta node at cluster-meta-node-02:8091
Searching for meta node on cluster-meta-node-03:8091...
Searching for data node on cluster-meta-node-03:8088...

Successfully joined cluster

  * Added meta node 18 at cluster-meta-node-03:8091
  * No data node added.  Run with -v to see more information

The command joins the meta node running at cluster-meta-node-03:8091 to an existing cluster. The existing cluster includes the meta node running at cluster-meta-node-02:8091. The system doesn’t join a data node to the cluster because it doesn’t find a data node at cluster-meta-node-03:8088.

Example 4: Join a meta node to an existing cluster and show detailed information about the join

$ influxd-ctl join -v meta-node-02:8091

Joining meta node at meta-node-02:8091
Searching for meta node on meta-node-03:8091...
Searching for data node on data-node-03:8088...

No data node found on data-node-03:8091!

  If a data node is running on this host,
  you may need to add it manually using the following command:

  influxd-ctl -bind meta-node-02:8091 add-data <dataAddr:port>

  Common problems:

    * The influxd process is using a non-standard port (default 8088)
    * The influxd process it not running.  Check the logs for startup errors

Successfully joined cluster

  * Added meta node 18 at meta-node-03:8091
  * No data node added.  Run with -v to see more information

The command joins the meta node running at cluster-meta-node-03:8091 to an existing cluster. The existing cluster includes the meta node running at cluster-meta-node-02:8091. The -v option prints detailed information about the join.

kill-copy-shard

Aborts an in-progress copy-shard command.

kill-copy-shard <data-node-source-TCP-address> <data-node-destination-TCP-address> <shard-id>
Examples


Example 1: Stop an in-progress copy-shard command

$ influxd-ctl kill-copy-shard cluster-data-node-02:8088 cluster-data-node-03:8088 39

Killed shard copy 39 from cluster-data-node-02:8088 to cluster-data-node-03:8088

The command aborts the copy-shard command that was copying shard 39 from cluster-data-node-02:8088 to cluster-data-node-03:8088.

leave

Removes a meta node and/or data node from the cluster. Use leave instead of the remove-meta and remove-data arguments if you set up your InfluxDB Enterprise cluster with the QuickStart Installation process.

The leave argument is destructive; it erases all metastore information from meta nodes and all data from data nodes. Use leave only if you want to permanently remove a node from a cluster.

leave [-y]

Options:

-y    Assume yes (y) to all prompts.

Examples


Example 1: Remove a meta and data node from a cluster

$ influxd-ctl leave

Searching for data node on cluster-node-03:8088...
Remove data node cluster-node-03:8088 from the cluster [y/N]: y
Removed cluster-node-03:8088 from the cluster
Searching for meta node on cluster-node-03:8091...
Remove meta node cluster-node-03:8091 from the cluster [y/N]: y

Successfully left cluster

  * Removed data node cluster-node-03:8088 from cluster
  * Removed meta node cluster-node-03:8091 from cluster

The command removes the meta node running at cluster-node-03:8091 and the data node running at cluster-node-03:8088 from an existing cluster. Here, we respond yes (y) to the two prompts that ask if we’d like to remove the data node and if we’d like to remove the meta node from the cluster.

Example 2: Remove a meta and data node from a cluster and assume yes to all prompts

$ influxd-ctl leave -y

Searching for data node on cluster-node-03:8088...
Removed cluster-node-03:8088 from the cluster
Searching for meta node on cluster-node-03:8091...

Successfully left cluster

  * Removed data node cluster-node-03:8088 from cluster
  * Removed meta node cluster-node-03:8091 from cluster

The command removes the meta node running at cluster-node-03:8091 and the data node running at cluster-node-03:8088 from an existing cluster. Because we specify the -y option, the system assumes that we’d like to remove both the data node and meta node from the cluster and does not prompt us for responses.

Example 3: Remove a meta node from a cluster

$ influxd-ctl leave

Searching for data node on cluster-meta-node-03:8088...
  * No data node found.
Searching for meta node on cluster-meta-node-03:8091...
Remove meta node cluster-meta-node-03:8091 from the cluster [y/N]: y

Successfully left cluster

  * No data node removed from cluster
  * Removed meta node cluster-meta-node-03:8091 from cluster

The command removes the meta node running at cluster-meta-node-03:8091 from an existing cluster. The system doesn’t remove a data node from the cluster because it doesn’t find a data node running at cluster-meta-node-03:8088.

remove-data

Removes a data node from a cluster. Use remove-data instead of the leave argument if you set up your InfluxDB Enterprise cluster with the Production Installation process.

The remove-data argument is destructive; it erases all data from the specified data node. Use remove-data only if you want to permanently remove a data node from a cluster.

remove-data [-force] <data-node-TCP-bind-address>

Options:

-force    Forces the removal of the data node. Use -force if the data node process is not running.

Examples


Example 1: Remove a data node from a cluster

~# influxd-ctl remove-data cluster-data-node-03:8088
Removed data node at cluster-data-node-03:8088

The command removes a data node running at cluster-data-node-03:8088 from an existing cluster.

remove-meta

Removes a meta node from the cluster. Use remove-meta instead of the leave argument if you set up your InfluxDB Enterprise cluster with the Production Installation process.

The remove-meta argument is destructive; it erases all metastore information from the specified meta node. Use remove-meta only if you want to permanently remove a meta node from a cluster.

remove-meta [-force | -tcpAddr <meta-node-TCP-bind_address> | -y] <meta-node-HTTP-bind-address>

Options:

-force    Forces the removal of the meta node. Use -force if the meta node process if not running, and the node is not reachable and unrecoverable. If a meta node restarts after being -force removed, it may interfere with the cluster. This options requires the -tcpAddr option.

-tcpAddr <meta-node-TCP-bind_address>    The TCP address of the meta node to remove from the cluster. Use this option with the -force option.

-y    Assumes Yes to all prompts.

Examples


Example 1: Remove a meta node from a cluster

$ influxd-ctl remove-meta cluster-meta-node-02:8091

Remove cluster-meta-node-02:8091 from the cluster [y/N]: y
y

Removed meta node at cluster-meta-node-02:8091

The command removes the meta node at cluster-meta-node-02:8091 from an existing cluster. In the example, we respond yes (y) to the prompt that asks if we’d like to remove the meta node from the cluster.

Example 2: Force remove an unresponsive meta node from a cluster

$ influxd-ctl remove-meta -force -tcpAddr cluster-meta-node-02:8089 cluster-meta-node-02:8091

Force remove cluster-meta-node-02:8091 from the cluster [y/N]:y
y

Removed meta node at cluster-meta-node-02:8091

The command force removes the meta node running at the TCP address cluster-meta-node-02:8089 and HTTP address cluster-meta-node-02:8091 from the cluster. In the example, we respond yes (y) to the prompt that asks if we’d like to force remove the meta node from the cluster. Note that if the meta node at cluster-meta-node-02:8091 restarts, it may interfere with the cluster. Only perform a force removal of a meta node if the node is not reachable and unrecoverable.

remove-shard

Removes a shard from a data node. Removing a shard is an irrecoverable, destructive action; please be cautious with this command.

remove-shard <data-node-source-TCP-address> <shard-id>

Resources: Cluster Rebalance

Examples


Example 1:

~# influxd-ctl remove-shard cluster-data-node-02:8088 31

Removed shard 31 from cluster-data-node-02:8088

The command removes shard 31 from the data node running at cluster-data-node-02:8088.

restore

Restore a backup to an existing cluster or a new cluster. Note that he existing cluster must contain no data in the databases affected by the restore. Restore supports both full backups and incremental backups; the syntax for a restore differs depending on the backup type.

restore [ -db <database> | -full | -list | -newdb <new-database> | -newrf <int> | -newrp <retention-policy> | -rp <retention policy | shard <shard-id> ] ( <path-to-backup-manifest-file> | <path-to-backup-directory> )

The restore command must specify either the path-to-backup-manifest-file or the path-to-backup-directory. If the restore uses the -full option, specify the path-to-backup-manifest-file. If the restore doesn’t use the -full option, specify the path-to-backup-directory.

Options:

-db <string>    The name of the single database to restore.

-full    Restore a backup that was created with the -full flag. A restore command with the -full flag requires the path-to-backup-manifest-file.

-list    Show the contents of the backup.

-newdb <string>    The name of the new database to restore to (must specify with -db).

-newrf <int>    The new replication factor to restore to (this is capped to the number of data nodes in the cluster).

-newrp <string>    The name of the new retention policy to restore to (must specify with -rp).

-rp <string>    The name of the single retention policy to restore.

-shard <unit>    The shard ID to restore.

Resources: Backup and Restore

Examples


Example 1: Restore from an incremental backup

$ influxd-ctl restore my-incremental-backup/

Using backup directory: my-incremental-backup/
Using meta backup: 20170130T231333Z.meta
Restoring meta data... Done. Restored in 21.373019ms, 1 shards mapped
Restoring db telegraf, rp autogen, shard 2 to shard 2...
Copying data to <hostname>:8088... Copying data to <hostname>:8088... Done. Restored shard 2 into shard 2 in 61.046571ms, 588800 bytes transferred
Restored from my-incremental-backup/ in 83.892591ms, transferred 588800 bytes

The command restores an incremental backup stored in the my-incremental-backup/ directory.

Example 2: Restore from a full backup

$ influxd-ctl restore -full my-full-backup/20170131T020341Z.manifest

Using manifest: my-full-backup/20170131T020341Z.manifest
Restoring meta data... Done. Restored in 9.585639ms, 1 shards mapped
Restoring db telegraf, rp autogen, shard 2 to shard 2...
Copying data to <hostname>:8088... Copying data to <hostname>:8088... Done. Restored shard 2 into shard 2 in 48.095082ms, 569344 bytes transferred
Restored from my-full-backup in 58.58301ms, transferred 569344 bytes

The command restores a full backup that includes the manifest file at my-full-backup/20170131T020341Z.manifest.

show

Shows all meta nodes and data nodes that are part of the cluster. The output includes the InfluxDB Enterprise version number.

show
Examples


Example 1: Show all meta and data nodes in a cluster

$ influxd-ctl show

Data Nodes
==========
ID	 TCP Address		        Version
2   cluster-node-01:8088	1.3.x-c1.3.x
4   cluster-node-02:8088	1.3.x-c1.3.x

Meta Nodes
==========
TCP Address		        Version
cluster-node-01:8091	1.3.x-c1.3.x
cluster-node-02:8091	1.3.x-c1.3.x
cluster-node-03:8091	1.3.x-c1.3.x

The output shows that the cluster includes three meta nodes and two data nodes. Every node is using InfluxDB Enterprise version 1.3.x-c1.3.x.

show-shards

Shows the cluster’s existing shards, including the shard’s id, database, retention policy, desired number of copies, shard group, start time, end time, expiry time and data node owners.

show-shards
Examples


Example 1: Show the existing shards in a cluster

$ influxd-ctl show-shards

Shards
==========
ID  Database             Retention Policy  Desired Replicas  Shard Group  Start                 End                              Expires               Owners
51  telegraf             autogen           2                 37           2017-03-13T00:00:00Z  2017-03-20T00:00:00Z                                   [{26 cluster-data-node-01:8088} {33 cluster-data-node-03:8088}]
52  telegraf             autogen           2                 37           2017-03-13T00:00:00Z  2017-03-20T00:00:00Z                                   [{5 cluster-data-node-02:8088} {26 cluster-data-node-01:8088}]

The output shows that there are two shards in the cluster. The first shard has an id of 51 and it’s in the telegraf database and the autogen retention policy. The desired number of copies for shard 51 is 2 and it belongs to shard group 37. The data in shard 51 cover the time range between 2017-03-13T00:00:00Z and 2017-03-20T00:00:00Z, and the shard has no expiry time; telegraf’s autogen retention policy has an infinite duration so the system never removes shard 51. Finally, shard 51 appears on two data nodes: cluster-data-node-01:8088 and cluster-data-node-03:8088.

update-data

Updates a data node’s address in the meta store.

update-data <data-node-old-TCP-bind-address> <data-node-new-TCP-bind-address>
Examples


Example 1: Update a data node’s hostname

$ influxd-ctl update-data cluster-node-01:8088 cluster-data-node-01:8088

updated data node 26 to cluster-data-node-01:8088

The command updates the address for data node 26 from cluster-node-01:8088 to cluster-data-node-01:8088.

token

Generates a signed JSON Web Token (JWT) token. The token argument only works when using JWT authentication in the cluster and when using the -auth-type jwt and -secret <shared-secret> flags.

token [-exp <duration>]

Options:

-exp    Determines the time after which the token expires. By default, the token expires after one minute.

Examples


Example 1: Create a signed JWT token

$ influxd-ctl -auth-type jwt -secret oatclusters token

hereistokenisitgoodandsoareyoufriend.timingisaficklefriendbutwherewouldwebewithoutit.timingthentimeseriesgood-wevemadetheleap-nowletsgetdownanddataandqueryallourheartsout

The command returns a signed JWT token.

Example 2: Attempt to create a signed JWT token with basic authentication

$ influxd-ctl -auth-type basic -user admini -pwd mouse token

token: tokens can only be created when using bearer authentication

The command returns an error because the command doesn’t use JWT authentication.

truncate-shards

Truncates hot shards, that is, shards that cover the time range that includes the current time (now()). truncate-shards creates a new shard and the system writes all new points to that shard.

truncate-shards [-delay <duration>]

Options:

-delay <duration>    Determines when to truncate shards after now(). By default, the tool sets the delay to one minute. The duration is an integer followed by a duration unit.

Resources: Cluster Rebalance

Examples


Example 1: Truncate shards with the default delay time

$ influxd-ctl truncate-shards

Truncated shards.

$ influxd-ctl show-shards

Shards
==========
ID  Database             Retention Policy  Desired Replicas  Shard Group  Start                 End                              Expires               Owners
51  telegraf             autogen           2                 37           2017-03-13T00:00:00Z  2017-03-13T20:40:15.753443255Z*                        [{26 cluster-data-node-01:8088} {33 cluster-data-node-03:8088}]
54  telegraf             autogen           2                 38           2017-03-13T00:00:00Z  2017-03-20T00:00:00Z                                   [{26 cluster-data-node-01:8088} {33 cluster-data-node-03:8088}]

After running influxd-ctl truncate-shards and waiting one minute, the output of the influxd-ctl show-shards command shows that the system truncated shard 51 (truncated shards have an asterix (*) on the timestamp in the End column) and created the new shard with the id 54.

Example 2: Truncate shards with a user-provided delay timestamp

$ influxd-ctl truncate-shards -delay 3m

Truncated shards.

$ influxd-ctl show-shards

Shards
==========
ID  Database             Retention Policy  Desired Replicas  Shard Group  Start                 End                              Expires               Owners
54  telegraf             autogen           2                 38           2017-03-13T00:00:00Z  2017-03-13T20:59:14.665827038Z*                        [{26 cluster-data-node-01:8088} {33 cluster-data-node-03:8088}]
58  telegraf             autogen           2                 40           2017-03-13T00:00:00Z  2017-03-20T00:00:00Z                                   [{26 cluster-data-node-01:8088} {33 cluster-data-node-03:8088}]

After running influxd-ctl truncate-shards and waiting three minutes, the output of the influxd-ctl show-shards command shows that the system truncated shard 54 (truncated shards have an asterix (*) on the timestamp in the End column) and created the new shard with the id 58.

influx

The influx tool, also known as the Command Line Interface (CLI), is available on all data nodes. Use influx to write data to your cluster, query data interactively, and view query output in different formats.

The complete description of the influx tool is available in the InfluxDB OSS documentation.