Configuration

Configuration

Many options are available to configure a MeiliSearch instance. Each of these options is added on MeiliSearch instance launch.

Options can be either communicated through environment variables or command line options. If both are provided for the same option, the command line option value is kept.

Passing arguments via the command line

Options are added at launch.

$ ./meilisearch --db-path ./meilifiles --http-addr '127.0.0.1:7700'
Server is listening on: http://127.0.0.1:7700

Passing arguments via the environment variables

The format of the environment variables is identical to the command line options with the exception that it is uppercased and MEILI_ is prepended.

$ export MEILI_DB_PATH=./meilifiles
$ export MEILI_HTTP_ADDR=127.0.0.1:7700
$ ./meilisearch
Server is listening on: http://127.0.0.1:7700

Options

General

Advanced

Database path

Environment variable: MEILI_DB_PATH
CLI option: --db-path

Defines the location for the database files.

Default value: "./data.ms"

HTTP address & port binding

Environment variable: MEILI_HTTP_ADDR
CLI option: --http-addr

The address the HTTP server will listen on.

Default value: "127.0.0.1:7700"

Master Key

Environment variable: MEILI_MASTER_KEY
CLI option: --master-key

The master key allowing you to do everything on the server. If no master key is provided all routes will be accessible without keys. This is only possible if you are in development environment. An error is thrown if you try to start MeiliSearch without any master key when the environment is set to production.

Learn more about the permission and authentication in this guide.

Default value: None

Analytics

Environment variable: MEILI_NO_ANALYTICS
CLI option: --no-analytics

Deactivates analytics.

Analytics allow us to know how many users are using MeiliSearch and the following:

The application version.
The environment: development/production.
The number of days since the first start: Segment development/production.
The user email: (Optional) If the user wants to have alerts.
The server provider: (Optional).
The database size.
The last update time.
The number of updates.
The number of documents per index.

Default value: false

Environment

Environment variable: MEILI_ENV
CLI option: --env

By default, MeiliSearch runs in development mode.

production: the master key is mandatory.
development: the master key is optional, and logs are output in “info” mode (console output).

If the server is running in development mode more logs will be displayed, and the master key can be avoided which implies that there is no security on the updates routes.
This is useful to debug when integrating the engine with another service.
In production mode, the web interface is disabled.

Default value: development

Payload Limit Size

Environment variable: MEILI_HTTP_PAYLOAD_SIZE_LIMIT
CLI option: --http-payload-size-limit

The maximum size, in bytes, of accepted JSON payloads.

Default value: 104857600 (+=100MB)

SSL Authentication Path

Environment variable: MEILI_SSL_AUTH_PATH
CLI option: --ssl-auth-path

Enable client authentication, and accept certificates signed by those roots provided in CERTFILE

Default value: None

SSL Certificates Path

Environment variable: MEILI_SSL_CERT_PATH
CLI option: --ssl-cert-path

Read server certificates from CERTFILE. This should contain PEM-format certificates in the right order (the first certificate should certify KEYFILE, the last should be a root CA).

Default value: None

SSL Key Path

Environment variable: MEILI_SSL_KEY_PATH
CLI option: --ssl-key-path

Read private key from KEYFILE. This should be a RSA private key or PKCS8-encoded private key, in PEM format.

Default value: None

SSL OCSP path

Environment variable: MEILI_SSL_OCSP_PATH
CLI option: --ssl-ocsp-path

Read DER-encoded OCSP response from OCSPFILE and staple to certificate. Optional.

Default value: None

SSL Require Auth

Environment variable: MEILI_SSL_REQUIRE_AUTH
CLI option: --ssl-require-auth

Send a fatal alert if the client does not complete client authentication.

Default value: None

SSL Resumption

Environment variable: MEILI_SSL_RESUMPTION
CLI option: --ssl-resumption

SSL support session resumption.

Default value: None

SSL Tickets

Environment variable: MEILI_SSL_TICKETS
CLI option: --ssl-tickets

SSL support tickets.

Default value: None

Max MDB Size

Environment variable: MEILI_MAX_MDB_SIZE
CLI option: --max-mdb-size

The maximum size, in bytes, of the main database. The main database stores the processed data.

The size must be a modulo value of your OS PAGE_SIZE otherwise it will throw an error.
You can find out about the PAGE_SIZE with the following command:

getconf PAGE_SIZE

Depending on the OS, it is either the size that will be allocated on launch or the maximum size the database can attain.

On UNIX it is the maximum size.
On Windows it is a fixed size that will be allocated on launch.
Because this allocates 100Gb on MeiliSearch launch, a Windows user can use this option to decrease the size of the database.

To know more about storage in MeiliSearch look at this guide

Default value: 107374182400 (100 GiB)

Max UDB Size

Environment variable: MEILI_MAX_UDB_SIZE
CLI option: --max-udb-size

The maximum size, in bytes, of the update database. The update database stores the pending updates.

The size must be a modulo value of your OS PAGE_SIZE otherwise it will throw an error.
You can find out about the PAGE_SIZE with the following command:

getconf PAGE_SIZE

Depending on the OS, it is either the size that will be allocated on launch or the maximum size the database can attain.

On UNIX it is the maximum size.
On Windows it is a fixed size that will be allocated on launch.
Because this allocates 100Gb on MeiliSearch launch, a Windows user can use this option to decrease the size of the database.

To know more about storage in MeiliSearch look at this guide

Default value: 107374182400 (100 GiB)

Disable Sentry

Environment variable: MEILI_NO_SENTRY
CLI option: --no-sentry

We use Sentry (opens new window) to get bug reports and diagnostics, and improve MeiliSearch experience. You can disable it at any moment by launching MeiliSearch with the argument above.

Schedule Snapshot Creation

Environment variable: MEILI_SCHEDULE_SNAPSHOT
CLI option: --schedule-snapshot

Activates scheduled snapshots.

If this command is not added or its value is false snapshotting is deactivated.

Snapshot Destination

Environment variable: MEILI_SNAPSHOT_DIR
CLI option: --snapshot-dir

The directory path where MeiliSearch will create snapshots.

Default value: snapshots/

Snapshot Interval

Environment variable: MEILI_SNAPSHOT_INTERVAL_SEC
CLI option: --snapshot-interval-sec

Defines the time gap in seconds between each snapshot creation.

Default value: 86400 (1 day)

Import Snapshot

CLI option: --import-snapshot

The path of the snapshot file to import.

This command will stop the process if:

A database already exists
No snapshot exists in the given path.

If this command is not called, no snapshot will be imported.

Ignore Missing Snapshot

CLI option: --ignore-missing-snapshot

The engine ignores missing snapshots and does not throw an error in this case.

Requires --import-snapshot to be defined.

Ignore Snapshot if DB Exists

CLI option: --ignore-snapshot-if-db-exists

If a database already exists, MeiliSearch will attempt to launch using that database instead of importing a snapshot. No error is thrown in this case.

Requires --import-snapshot to be defined.

Dumps Destination

Environment variable: MEILI_DUMPS_DIR
CLI option: --dumps-dir

Path of the directory where dumps will be created if the dump route is called.

Default value: dumps/

Import Dump

CLI option: --import-dump

Import a dump from the specified path. Must be a .dump file.

As the data contained in the dump needs to be indexed, the process will take an amount of time corresponding to the size of the dump. Only when the import is complete and successful will the MeiliSearch server start.

Dump Batch Size

Environment variable: MEILI_DUMP_BATCH_SIZE
CLI option: --dump-batch-size

Sets the batch size used in the dump importation process. This number corresponds to the maximum number of documents indexed in each batch. A larger value will take less time but use more memory.

If a dump import process is killed, this means that you do not have enough RAM. Consider reducing your batch size.

If you find that a dump import process is too slow and you have a lot of RAM to spare, consider increasing the batch size, as it will accelerate the indexation. However, if this leads to the dump process failing, you’ve gone too far and run out of memory. In this case, you should decrease the batch size until you find the right balance between speed and memory overhead.

Example
Imagine you set --dump-batch-size 1000 and your dump contains 2600 documents. Instead of indexing all 2600 docs in one go, the engine will :

Index documents 0 -> 999 (1000 docs)
Index documents 1000 -> 1999 (1000 docs)
Index documents 2000 -> 2599 (600 docs)

Default value: 1024