Skip to content

Official Microsoft Outlook Connector Package for Elastic Enterprise Search, App Search and Workplace Search

License

Notifications You must be signed in to change notification settings

elastic/enterprise-search-microsoft-outlook-connector

Repository files navigation

Elastic Enterprise Search | Elastic Workplace Search


For new users, we recommend using our Elasticsearch native tools, rather than the standalone Workplace Search product. See this blog post for more information about upgrading your internal knowledge search, to make it an amazing experience for your users!

We recommend using the new Elastic Outlook connector to ingest your content from Outlook in regular Elasticsearch indices.


Workplace Search: Microsoft Outlook connector package

Use this Elastic Enterprise Search Microsoft Outlook connector package to deploy and run a Microsoft Outlook content source on your own infrastructure. The connector package extracts and syncs data from Microsoft Exchange Outlook and Microsoft Office365 Outlook application. The data is indexed into a Workplace Search content source within an Elastic deployment.

⚠️ This connector package is a beta feature. Beta features are subject to change and are not covered by the support SLA of generally available (GA) features. Elastic plans to promote this feature to GA in a future release.

ℹ️ This connector package requires a compatible Elastic subscription level. Refer to the Elastic subscriptions pages for Elastic Cloud and self-managed deployments.

Table of contents:

Setup and basic usage

Complete the following steps to deploy and run the connector:

  1. Gather Microsoft Outlook details
  2. Gather Elastic details
  3. Create a Workplace Search API key
  4. Create a Workplace Search content source
  5. Choose connector infrastructure and satisfy dependencies
  6. Install the connector
  7. Configure the connector
  8. Test the connection
  9. Sync data
  10. Log errors and exceptions
  11. Schedule recurring syncs

The steps above are relevant to all users. Some users may require additional features. These are covered in the following sections:

Gather Microsoft Outlook details

First identify whether you are using Microsoft Exchange Outlook or Microsoft Office365 Outlook, and configure the parameter connector_platform_type. You'll need to gather different details depending on which service you are using.

For Microsoft Exchange Outlook, collect the following information:

  • The IP address or Hostname of the Microsoft Exchange active directory host.
  • The IP address or Hostname of the Microsoft Exchange Server host.
  • The username the connector will use to log in to Microsoft Exchange Outlook.
  • The password the connector will use to log in to Microsoft Exchange Outlook.
  • The domain of the Microsoft Exchange Server host.
  • The path of the SSL certificate if Microsoft Exchange Server host is secured.

ℹ️ The username and password must be the admin account for the Microsoft Exchange server.

For Microsoft Office365 Outlook, collect the following information:

  • The client id, tenant id and client secret will be used to generate access tokens and fetch users details.

Later, you will configure the connector with these values.

Some connector features require additional details. Review the following documentation if you plan to use these features:

Gather Elastic details

First, ensure your Elastic deployment is compatible with the Microsoft Outlook connector package.

Next, determine the Enterprise Search base URL for your Elastic deployment.

Later, you will configure the connector with this value.

You also need a Workplace Search API key and a Workplace Search content source ID. You will create those in the following sections.

If you plan to use document-level permissions, you will also need user identity information. See Use document-level permissions (DLP) for details.

Create a Workplace Search API key

Each Microsoft Outlook Server connector authorizes its connection to Elastic using a Workplace Search API key.

Create an API key within Kibana. See Workplace Search API keys.

Create a Workplace Search content source

Each Microsoft Outlook Server connector syncs data from Microsoft Outlook Server into a Workplace Search content source.

Create a content source within Kibana:

  1. Navigate to Enterprise Search → Workplace Search → Sources → Add Source → Custom API Source.
  2. Name your Content Source, (e.g. Microsoft Outlook Connector).
  3. Choose Configure Microsoft Outlook Connector.

For more details please refer Elastic Documentation for creating a custom API source.

Record the ID of the new content source. This value is labeled Source Identifier within Kibana. Later, you will configure the connector with this value.

Alternatively, you can use the connector’s bootstrap command to create the content source. See bootstrap command.

Choose connector infrastructure and satisfy dependencies

After you’ve prepared the two services, you are ready to connect them.

Provision a Windows, MacOS, or Linux server for your Microsoft Outlook connectors.

The infrastructure must provide the necessary runtime dependencies. See Runtime dependencies.

Clone or copy the contents of this repository to your infrastructure.

Install the connector

After you’ve provisioned infrastructure and copied the package, use the provided make target to install the connector:

make install_package

This command runs as the current user and installs the connector and its dependencies. Note: By Default, the package installed supports Enterprise Search version 8.0 or above. In order to use the connector for older versions of Enterprise Search(less than version 8.0) use the ES_VERSION_V8 argument while running the make install_package or make install_locally command:

make install_package ES_VERSION_V8=no

ℹ️ Within a Windows environment, first install make:

Note: If you are running the connector on Windows, please ensure Microsoft Visual C++ 14.0 or greater is installed.

winget install -e --id GnuWin32.Make

Next, ensure the ees_microsoft_outlook executable is on your PATH. For example, on macOS:

export PATH=/Users/$USER/Library/Python/3.8/bin:$PATH

The following table provides the installation location for each operating system (note Python version 3.8):

Operating system Installation location
Linux ./local/bin
macOS /Users/<user_name>/Library/Python/3.8/bin
Windows \Users\<user_name>\AppData\Roaming\Python\Python38\Scripts

Configure the connector

You must configure the connector to provide the information necessary to communicate with each service. You can provide additional configuration to customize the connector for your needs.

Create a YAML configuration file at any pathname. Later, you will include the -c option when running commands to specify the pathname to this configuration file.

Alternatively, in Linux environments only, locate the default configuration file created during installation. The file is named microsoft_outlook_connector.yml and is located within the config subdirectory where the package files were installed. See Install the connector for a listing of installation locations by operating system. When you use the default configuration file, you do not need to include the -c option when running commands.

After you’ve located or created the configuration file, populate each of the configuration settings. Refer to the settings reference. You must provide a value for all required settings.

Use the additional settings to customize the connection and manage features such as document-level permissions. See:

Test the connection

After you’ve configured the connector, you can test the connection between Elastic and Microsoft Outlook. Use the following make target to test the connection:

make test_connectivity

Sync data

After you’ve confirmed the connection between the two services, you are ready to sync data from Microsoft Outlook to Elastic.

The following table lists the available sync operations, as well as the commands to perform the operations.

Operation Command
Incremental sync incremental-sync
Full sync full-sync
Deletion sync deletion-sync
Permission sync permission-sync

Begin syncing with an incremental sync. This operation begins extracting and syncing content from Microsoft Outlook to Elastic. If desired, customize extraction and syncing for your use case.

Review the additional sync operations to learn about the different types of syncs. Additional configuration is required to use document-level permissions.

You can use the command line interface to run sync operations on demand, but you will likely want to schedule recurring syncs.

Log errors and exceptions

The various sync commands write logs to standard output and standard error.

To persist logs, redirect standard output and standard error to a file. For example:

ees_microsoft_outlook -c ~/config.yml incremental-sync >>~/incremental-sync.log 2>&1

You can use these log files to implement your own monitoring and alerting solution.

Configure the log level using the log_level setting.

Schedule recurring syncs

Use a job scheduler, such as cron, to run the various sync commands as recurring syncs.

The following is an example crontab file in linux:

PATH=/home/<user_name>/.local/bin
0 */2 * * * ees_microsoft_outlook -c ~/config.yml incremental-sync >>~/incremental-sync.log 2>&1
0 0 */2 * * ees_microsoft_outlook -c ~/config.yml full-sync >>~/full-sync.log 2>&1
0 * * * * ees_microsoft_outlook -c ~/config.yml deletion-sync >>~/deletion-sync.log 2>&1
*/5 * * * * ees_microsoft_outlook -c ~/config.yml permission-sync >>~/permission-sync.log 2>&1

This example redirects standard output and standard error to files, as explained here: Log errors and exceptions.

Use this example to create your own crontab file. Manually add the file to your crontab using crontab -e. Or, if your system supports cron.d, copy or symlink the file into /etc/cron.d/.

⚠️ Note: It's possible that scheduled jobs may overlap. To avoid multiple crons running concurrently, you can use flock with cron to manage locks. The flock command is part of the util-linux package. You can install it with yum install util-linux or sudo apt-get install -y util-linux. Using flock ensures the next scheduled cron runs only after the current one has completed execution.

Let's consider an example of running incremental-sync as a cron job with flock:

0 */2 * * * /usr/bin/flock -w 0 /var/cron.lock ees_microsoft_outlook -c ~/config.yml incremental-sync >>~/incremental-sync.log 2>&1

Note: If the flock is added for multiple commands in crontab, make sure you mention different lock names(eg: /var/cron_indexing.lock in the above example) for each job else the execution of one command will prevent other command to execute.

Troubleshooting

To troubleshoot an issue, first view your logged errors and exceptions.

Use the following sections to help troubleshoot further:

If you need assistance, use the Elastic community forums or Elastic support:

Troubleshoot extraction

The following sections provide solutions for content extraction issues.

Issues extracting content from attachments

The connector uses the Tika module for parsing file contents from attachments. Tika-python uses Apache Tika REST server. To use this library, you need to have Java 7+ installed on your system as tika-python starts up the Tika REST server in the background.

At times, the TIKA server fails to start hence content extraction from attachments may fail. To avoid this, make sure Tika is running in the background.

Issues extracting content from images

Tika Server also detects contents from images by automatically calling Tesseract OCR. To allow Tika to also extract content from images, you need to make sure tesseract is on your path and then restart tika-server in the background (if it is already running). For example, on a Unix-like system, try:

ps aux | grep tika | grep server # find PID
kill -9 <PID>

To allow Tika to extract content from images, you need to manually install Tesseract OCR.

Advanced usage

The following sections cover additional features that are not covered by the basic usage described above.

After you’ve set up your first connection, you may want to further customize that connection or scale to multiple connections.

Customize extraction and syncing

By default, each connection syncs all supported Microsoft Outlook data across all Microsoft Outlook applications.

You can also customize which objects are synced, and which fields are included and excluded for each object. Configure the setting objects.

Finally, you can set custom timestamps to control which objects are synced, based on their created or modified timestamps. Configure the following settings:

Troubleshoot Access Token Generation

The following sections provide the solution for issues related to access token generation.

Disable Multi-factor Authentication

  1. Go to Microsoft Azure AD -> Properties.
  2. Go to Manage Security defaults, disable the security, and save the changes.
  3. Go to Users and create a new user with global permissions from assignees roles.
  4. Go to Microsoft Outlook Azure AD conditional access and create a new policy:
  • Name: Name of the policy
  • Users or workload identities: include "allusers" and exclude the newly created users (this step will disable MFA for all excluded users).
  • Cloud apps or actions: include "All cloud apps"
  • Grant: select "grant access" with Require "multi-factor authentication" enabled and from multiple controls select "Require all the selected controls"

Add permissions to Microsoft Azure Platform

  1. Check the configuration file and verify all Microsoft Outlook settings configuration values are set correctly.
  2. If configuration values are set correctly, go to your application on Microsoft Azure Platform and verify all permissions are added as per the permission listed below and have the admin consent to each permission.
  • User.Read (Delegated)
  • User.Read.All (Delegated and Application)
  • full_access_as_app (Application)

Add permissions to Microsoft Exchange Server

  1. Check the configuration file and verify all Microsoft Outlook settings configuration values are set correctly.
  2. If configuration values are set correctly:
    1. Go to your application on Microsoft Exchange Server
    2. Verify permissions under admin roles
    3. Add Application Impersonation in the impersonation section
    4. Include admin user in member section.

Use document-level permissions (DLP)

Complete the following steps to use document-level permissions:

  1. Enable document-level permissions
  2. Map user identities
  3. Sync document-level permissions data

Enable document-level permissions

Within your configuration, enable document-level permissions using the following setting: enable_document_permission.

Map user identities

Copy to your server a CSV file that provides the mapping of user identities. The file must follow this format:

  • First column: Microsoft Outlook username
  • Second column: Elastic username

Then, configure the location of the CSV file using the following setting: connector.user_mapping.

Sync document-level permissions data

Sync document-level permissions data from Microsoft Outlook to Elastic.

The following sync operations include permissions data:

Sync this information continually to ensure correct permissions. See Schedule recurring syncs.

Connector reference

The following reference sections provide technical details:

Data extraction and syncing

Each Microsoft Outlook connector extracts and syncs the following data from Microsoft Outlook:

  • Mails
  • Calendar
  • Tasks
  • Contacts

The connector handles Microsoft Outlook pages comprised of various web parts, it extracts content from various document formats, and it performs optical character recognition (OCR) to extract content from images.

You can customize extraction and syncing per connector. See Customize extraction and syncing.

Sync operations

The following sections describe the various operations to sync data from Microsoft Outlook to Elastic.

Incremental sync

Syncs to Enterprise Search all supported Microsoft Outlook data created or modified since the previous incremental sync.

Perform this operation with the incremental-sync command.

Full sync

Syncs to Enterprise Search all supported Microsoft Outlook data created or modified since the configured start_time. Continues until the current time or the configured end_time.

Perform this operation with the full-sync command.

Deletion sync

Deletes from Enterprise Search all supported Microsoft Outlook data deleted since the previous deletion sync.

Perform this operation with the deletion-sync command.

Permission sync

Syncs to Enterprise Search all Microsoft Outlook document permissions since the previous permission sync.

When using document-level permissions (DLP), use this operation to sync all updates to users within Microsoft Outlook.

Perform this operation with the permission-sync command.

Command line interface (CLI)

Each Microsoft Outlook Server connector has the following command line interface (CLI):

ees_microsoft_outlook [-c <pathname>] <command>

-c option

The pathname of the configuration file to use for the given command.

ees_microsoft_outlook -c ~/config.yml full-sync

bootstrap command

Creates a Workplace Search content source with the given name. Outputs its ID.

ees_microsoft_outlook bootstrap --name 'Accounting documents' --user 'shay.banon'

See also Create a Workplace Search content source.

To use this command, you must configure the following settings:

And you must provide on the command line any of the following arguments that are required:

  • --name (required): The name of the Workplace Search content source to create.
  • --user (optional): The username of the Elastic user who will own the content source. If provided, the connector will prompt for a password. If omitted, the connector will use the configured API key to create the content source.

incremental-sync command

Performs an incremental sync operation.

full-sync command

Performs a full sync operation.

deletion-sync command

Performs a deletion sync operation.

permission-sync command

Performs a permission sync operation.

Configuration settings

Configure any of the following settings for a connector:

connector_platform_type (required)

The connector platform type. Allowed values for connector_platform_type are Microsoft Exchange and Office365.

connector_platform_type: 'Microsoft Exchange'

By default, it is set to Office365.

microsoft_exchange.active_directory_server (required for "Microsoft Exchange")

Microsoft Exchange active directory IP address or Hostname.

microsoft_exchange.active_directory_server: 0.0.0.0

microsoft_exchange.server (required for "Microsoft Exchange")

Microsoft Exchange Server IP address or Hostname.

microsoft_exchange.server: 0.0.0.0

microsoft_exchange.username (required for "Microsoft Exchange")

The username of the admin account for the Microsoft Exchange Server. See Gather Microsoft Outlook details.

microsoft_exchange.username: 'bill.gates'

microsoft_exchange.password (required for "Microsoft Exchange")

The password of the admin account for the Microsoft Exchange Server. See Gather Microsoft Outlook details.

microsoft_exchange.password: 'L,Ct%ddUvNTE5zk;GsDk^2w)(;,!aJ|Ip!?Oi'

microsoft_exchange.domain (required for "Microsoft Exchange")

The domain name for the Microsoft Exchange Server. See Gather Microsoft Outlook details.

microsoft_exchange.domain: 'abc.xyz'

microsoft_exchange.secure_connection (required for "Microsoft Exchange")

Validate the SSL certificate if host is secured. Specify Yes if host is secured and want to validate the SSL certificate, else No. See Gather Microsoft Outlook details.

microsoft_exchange.secure_connection: Yes

By default, it is set to Yes.

microsoft_exchange.certificate_path (required for "Microsoft Exchange")

The path of the SSL certificate to establish a secure connection with Microsoft Exchange server. See Gather Microsoft Outlook details.

microsoft_exchange.certificate_path: 'C:/Users/banon/microsoft_outlook_1/certificate.cer'

office365.client_id (required for "Office365")

Client ID of Microsoft Azure application. See Gather Microsoft Outlook details.

office365.client_id: 'a122dsad123334'

office365.tenant_id (required for "Office365")

Tenant ID of Microsoft Azure application. See Gather Microsoft Outlook details.

office365.tenant_id: 'a122dsad123334a122dsad123334a122dsad123334'

office365.client_secret (required for "Office365")

Client Secret of Microsoft Azure application. See Gather Microsoft Outlook details.

office365.client_secret: 'a122dsad123334a122dsad123334'

enterprise_search.api_key (required)

The Workplace Search API key. See Create a Workplace Search API key.

enterprise_search.api_key: 'zvksftxrudcitxa7ris4328b'

enterprise_search.source_id (required)

The ID of the Workplace Search content source. See Create a Workplace Search content source.

enterprise_search.source_id: '62461219647336183fc7652d'

enterprise_search.host_url (required)

The Enterprise Search base URL for your Elastic deployment. Note: To use Enterprise Search version 8 or above, port is required.

enterprise_search.host_url: 'https://my-deployment.ent.europe-west1.gcp.cloud.es.io:9243'

enable_document_permission

Whether the connector should sync document-level permissions (DLP) from Microsoft Outlook. By default, it is set to Yes i.e. by default the connector will try to sync document-level permissions.

enable_document_permission: Yes

objects

Specifies which Microsoft Outlook objects to sync to Enterprise Search, and for each object, which fields to include and exclude. When the include/exclude fields are empty, all fields are synced.

objects:
  mails:
    include_fields:
    exclude_fields:
  calendar:
    include_fields:
    exclude_fields:
  tasks:
    include_fields:
    exclude_fields:
  contacts:
    include_fields:
    exclude_fields:

start_time

A UTC timestamp the connector uses to determine which objects to extract and sync from Microsoft Outlook. Determines the starting point for a full sync. Supports the following time format YYYY-MM-DDTHH:MM:SSZ By default it is set to 180 days from current time.

start_time: 2022-04-01T04:44:16Z

end_time

A UTC timestamp the connector uses to determine which objects to extract and sync from Microsoft Outlook. Determines the stopping point for a full sync. Supports the following time format YYYY-MM-DDTHH:MM:SSZ By default it is set to current execution time.

end_time: 2022-04-01T04:44:16Z

log_level

The level or severity that determines the threshold for logging a message. One of the following values:

  • DEBUG
  • INFO (default)
  • WARNING
  • ERROR
log_level: 'INFO'

By default, it is set to INFO.

retry_count

The number of retries to perform when there is a server error. The connector applies an exponential backoff algorithm to retries.

retry_count: 3

By default, it is set to 3.

source_sync_sync_thread_count

The number of threads the connector will run in parallel to fetch documents from the Microsoft Outlook. By default, the connector uses 5 threads.

source_sync_sync_thread_count: 5

enterprise_search_sync_thread_count

The number of threads the connector will run in parallel for indexing documents to the Enterprise Search instance. By default, the connector uses 5 threads.

enterprise_search_sync_thread_count: 5

For a Linux distribution with at least 2 GB RAM and 4 vCPUs, you can increase the thread counts if the overall CPU and RAM are under utilized, i.e. below 60-70%.

connector.user_mapping

The pathname of the CSV file containing the user identity mappings for document-level permissions (DLP).

connector.user_mapping: 'C:/Users/banon/microsoft_outlook_1/identity_mappings.csv'

Enterprise Search compatibility

The Microsoft Outlook connector package is compatible with Elastic deployments that meet the following criteria:

  • Elastic Enterprise Search version greater than or equal to 7.13.0.
  • An Elastic subscription that supports this feature. Refer to the Elastic subscriptions pages for Elastic Cloud and self-managed deployments.

Runtime dependencies

Each Microsoft Outlook connector requires a runtime environment that satisfies the following dependencies:

  • Windows, MacOS, or Linux server. The connector has been tested with CentOS 7, MacOS Monterey v12.0.1, and Windows 10.
  • Python version 3.6 or later.
  • To extract content from images: Java version 7 or later, and tesseract command installed and added to PATH
  • To schedule recurring syncs: a job scheduler, such as cron

Connector Limitations

The following sections provide limitations of connector:

  • The connector can't fetch files larger than 10 MB due to a limitation with the exchange-lib module. These files will not be indexed into Workplace Search.
  • The exchange-lib module can be slow to return attachments from Mails, Calendars and Tasks. In some cases, attachments may be missing following an incremental sync. However, these missing attachments will be indexed in the next full-sync cycle.