The command-line interface of Mastodon is an executable file called `tootctl` residing in the `bin` directory within the Mastodon root directory. You must specify which environment you intend to use whenever you execute it by specifying the `RAILS_ENV` environment variable. Unless you are a developer working on a local machine, you need to use `RAILS_ENV=production`. If you are sure that you will never need another environment (for development, testing, or staging), you can add it to your `.bashrc` file for convenience, e.g.:
If so, you won’t need to specify it each time inline. Otherwise, calls to `tootctl` will usually go like this, assuming that the Mastodon code is checked out in `/home/mastodon/live`:
Erase this server from the federation by broadcasting account Delete activities to all known other servers. This allows a "clean exit" from running a Mastodon server, as it leaves next to no cache behind on other servers. This command is always interactive and requires confirmation twice.
No local data is actually deleted, because emptying the database or deleting the entire VPS is faster. If you run this command then continue to operate the instance anyway, then there will be a state mismatch that might lead to glitches and issues with federation.
{{<hintstyle="danger">}}
**Make sure you know exactly what you are doing before running this command.** This operation is NOT reversible, and it can take a long time. The server will be in a BROKEN STATE after this command finishes. A running Sidekiq process is required, so do not shut down the server until the queues are fully cleared.
: Define the new account's custom role by providing the `name` of that [Role]({{< relref "entities/Role" >}}). Default roles include `Owner`, `Admin`, and `Moderator` (case-sensitive).
: Define the existing account's custom role by providing the `name` of that [Role]({{< relref "entities/Role" >}}). Default roles include `Owner`, `Admin`, and `Moderator` (case-sensitive).
4.0.0 - `--role` no longer takes hard-coded `user`, `moderator`, or `admin` roles. Specify the name of the custom Role instead (case-sensitive). Remove the current role with `--remove-role`.
Request a backup for a user account with given USERNAME. The backup will be created in Sidekiq asynchronously, and the user will receive an email with a link to it once it's done.
Remove remote accounts that no longer exist. Queries every single remote account in the database to determine if it still exists on the origin server, and if it doesn't, then remove it from the database. Accounts that have had confirmed activity within the last week are excluded from the checks, in case the server is just down.
Merge two remote accounts into one. This is primarily meant to fix duplicates caused by other servers changing their domain. By default, this only works if the public key is the same, but this can be overridden.
Update hard-cached counters of TYPE by counting referenced records from scratch. It may take a very long time to finish, depending on the size of the database. Accounts will have their follower, following, and status counts refreshed. Statuses will have their reply, boost, and favourite counts refreshed.
Remove all accounts from a given DOMAIN without leaving behind any records. Unlike a suspension, if the DOMAIN still exists in the wild, it means the accounts could return if they are resolved again.
: Match domains in the actor URI rather than in the Webfinger address.
`--limited-federation-mode`
: Can be provided instead of DOMAIN. Instead of purging from a single domain, all accounts from domains that are not allow-listed will be removed from the database. Use this after enabling limited federation mode and defining your allow-list.
`--concurrency N`
: The number of workers to use for this task. Defaults to 5.
`--verbose`
: Print additional information while task is processing.
`--dry-run`
: Print expected results only, without performing any actions.
Crawl the known fediverse by using Mastodon REST API endpoints that expose all known peers, and collect statistics from those peers, as long as those peers support those API endpoints. When no START is given, the command uses the server's own database of known peers to seed the crawl. Returns total servers, total registered users, total active users in the last week, and total users joined in the last week.
: Do not include instances that you have suspended in the output. Also includes any subdomains.
`--concurrency N`
: The number of workers to use for this task. Defaults to N=50.
`--format FORMAT`
: Control how results are returned. `summary` will print a summary. `domains` will return a newline-delimited list of all discovered peers. `json` will dump aggregated raw data. Defaults to `summary`.
Imports custom emoji from a .tar.gz archive at a given path. The archive should contain PNG or GIF files no larger than 50KB, and the shortcode will be set equal to the filename minus the extension, with optional prefixes and/or suffixes.
Build home and list feeds for one or all users. Feeds will be built from the database and cached in-memory with Redis. Mastodon manages home feeds for active users automatically.
Fix corrupted database indexes that may have been caused due to changing collation rules. Deletes or merges duplicate accounts, statuses, emojis, etc. Mastodon has to be stopped to run this task, which will take a long time and may be destructive. This is useful if your database indexes are corrupted because of issues such as <https://wiki.postgresql.org/wiki/Locale_data_changes>.
: How old media attachments have to be before they are removed. In case of avatars and headers, how old the last webfinger request and update to the user has to be before they are removed. Defaults to 7.
: Instead of media attachments, remove locally cached copies of avatars and headers from other servers. Cannot be combined with `--remove-headers`.
`--remove-headers`
: Instead of media attachments, remove locally cached copies of headers from other servers. Cannot be combined with `--prune-profiles`.
`--include-follows`
: Override the default behavior of `--prune-profiles` and `--remove-headers` to remove locally cached copies of avatars (and headers) from other servers, irrespective of follow status (by default, they are only removed from accounts that are not followed by or following anyone locally). Can only be used with `--prune-profiles` or `--remove-headers`.
Scans for files that do not belong to existing media attachments, and remove them. Please mind that some storage providers charge for the necessary API requests to list objects. Also, this operation requires iterating over every single file individually, so it will be slow.
Refetch remote media attachments from other servers. You must specify the source of media attachments with either `--status`, `--account`, `--domain`, or `--days`. If an attachment already exists in the database, it will not be overwritten unless you use `--force`.
: How old media attachments have to be before they are removed. Defaults to 180. (NOTE: it is not recommended to delete preview cards within the last 14 days, because preview cards will not be refetched unless the link is reposted after 2 weeks from last time.)
`--concurrency N`
: The number of workers to use for this task. Defaults to N=5.
`--verbose`
: Print additional information while task is processing.
`--dry-run`
: Print expected results only, without performing any actions.
`--link`
: Only delete link-type preview cards; leave video and photo cards untouched.
Create or update an Elasticsearch index and populate it. If Elasticsearch is empty, this command will create the necessary indices and then import data from the database into those indices. This command will also upgrade indices if the underlying schema has been changed since the last run.
: Defaults to 100. A higher batch size can make Elasticsearch process records more quickly, with less load on the PostgreSQL database, but can increase memory pressure on the Elasticsearch nodes during indexing.
Remove unreferenced statuses from the database, such as statuses that came from relays or from users who are no longer followed by any local accounts, and have not been replied to or otherwise interacted with.
This is a computationally heavy procedure that creates extra database indices before commencing, and removes them afterward.
Upgrade the storage schema to store all non-local media resources in a top-level cache directory. WARNING: This is optional, and only for deployments made before v3.1.4. This command can incur massive object storage costs due to moving potentially terabytes of data.
