Skip to main content

File System Crawler Quick Start

This guide gets you from zero to a running file system crawl in under 5 minutes.

Built-in baseline, not a hard limit

The source protocols and examples in this guide represent built-in support in Norconex Crawler v4. They are practical defaults, not a fixed ceiling. Teams can extend crawler behavior with custom connectors, parser logic, and pipeline components for customer-specific requirements.

Windows users

Replace crawl-fs.sh with crawl-fs.bat and ./ with .\ in all commands below.

Step 1 — Create a config file​

Create my-fs-crawl.yaml with the following minimal configuration:

id: my-first-crawl
numThreads: 5
startReferences:
  - /path/to/your/documents
committers:
  - class: LogCommitter
    ignoreContent: true
    logLevel: INFO

This crawl will process every file under the given path and log each document (LogCommitter is built-in and ideal for testing before connecting a real backend).

Step 2 — Filter by file type (optional)​

To crawl only specific file extensions:

id: my-first-crawl
startReferences:
  - /path/to/your/documents
referenceFilters:
  - class: ExtensionReferenceFilter
    onMatch: INCLUDE
    extensions:
      - pdf
      - docx
      - xlsx
      - pptx
      - txt
      - html

Step 3 — Start the crawl​

./crawl-fs.sh start -config=my-fs-crawl.yaml

You'll see log output as files are fetched, filtered, and committed.

Docker alternative​

If you prefer running with Docker, mount your config and logs, then run:

docker run --rm \
  -v "${PWD}:/opt/norconex/crawler/configs" \
  -v "${PWD}/logs:/opt/norconex/crawler/logs" \
  -e COLLECTOR_CONFIG_FILE=my-fs-crawl.yaml \
  norconex/crawler-fs:latest
During the v4 beta

No latest tag is published until 4.0.0 ships. Use the current pre-release tag instead (for example norconex/crawler-fs:4.0.0-beta-1); the download page always shows it.

For Docker Compose examples and release-tag guidance, see Docker.

Step 4 — Stop and resume​

Stop the crawl at any time:

./crawl-fs.sh stop -config=my-fs-crawl.yaml

Norconex saves its state automatically. Resume exactly where you left off by running the same start command again:

./crawl-fs.sh start -config=my-fs-crawl.yaml

To clear the crawler state before your next run, you can issue this command:

./crawl-fs.sh clean -config=my-fs-crawl.yaml

Alternatively, you can combine "clean" with the start command:

./crawl-fs.sh start -clean -config=my-fs-crawl.yaml

Step 5 — Remote file systems​

The file system crawler supports remote protocols out of the box. A few examples:

If you are unsure which scheme or start reference format to use for a given fetcher, see FS Fetchers Quickstart.

SFTP:

id: my-first-crawl
startReferences:
  - sftp://fileserver.example.com/data/documents
credentials:
  username: user
  password: myPassword

WebDAV (Nextcloud, older versions of SharePoint):

id: my-first-crawl
startReferences:
  - https://sharepoint.example.com/sites/mysite/Shared Documents/
credentials:
  username: user
  password: myPassword

Apache HDFS:

id: my-first-crawl
startReferences:
  - webhdfs://namenode:9870/user/data/corpus

For Box, Google Drive, Egnyte, and M365-specific examples, use the dedicated fetcher reference pages so configuration examples stay canonical in one place:

CMIS is a standard, so one fetcher covers Alfresco, Nuxeo, OpenText, Documentum, IBM FileNet and any other compliant repository. See Content Sources for the full list.

Step 6 — Send to a real target​

Replace the LogCommitter with your actual destination. See the Integrations page for all available committers and their configuration.

When using ZIP distributions, external committers are downloaded separately as nx-committer-<name>-<version>.zip and their lib/*.jar files must be copied into the crawler lib/ directory. Built-in committers such as LogCommitter do not require this extra step.

CLI reference​

CommandDescription
start -config=<file>Start or resume a crawl
stop -config=<file>Gracefully stop a running crawl
clean -config=<file>Delete crawl state (forces a full recrawl)

Next steps​