Configuring web applications

Once you select the Web application type, a Web Configuration section appears with several tabs containing various configuration options:

Understanding URL patterns

The URL match patterns have the following structure:

<scheme>://<host><path>

The URL match patterns can contain wildcards, but a port number cannot be specified.

The system forces the use of a wildcard for the scheme *://<host><path>, so it always matches both HTTP and HTTPS.

The user administrating the business applications, views and enters patterns with a simplified structure:

<host><path>

Internally, all the software components must handle the full URL match patterns, including the scheme: 

<scheme>://<host><path>

Pattern validation logic

The application configuration module uses the following set of rules to verify the pattern before saving it:

1. The user must enter at least one URL match pattern.

2. Today, specifying a scheme like http://</code> or <code>https:// is not supported, it must be *://

3. If the user does not specify *:// the form automatically adds the *:// prefix.

4. If the user enters a URL with a port number, the form displays an error message.

5. The host component can start with *. followed by part of the fully qualified domain name (FQDN). Note that this pattern also matches a hostname that is exactly the expression entered after the *. wildcard.

6. The host component must end with a complete and valid domain name if it starts with a *. wildcard.

7. The host component can just contain a local hostname (for example, no domain name) if it does not start with a *. wildcard. This is to support internal web applications that do not use FQDN.

8. You must specify the path with at least /*. If the user does not enter a path, the form logic appends /* to match all paths.

9. The path must end with a * wildcard. If the path does not, the form logic appends * to the pattern.

10. The path cannot contain more than one * wildcard.

11. The pattern must not contain a query string (?...). If the user enters a URL with a query string, an error message appears.

12. The pattern must not contain a fragment (#...). If the user enters a URL with a fragment, an error message appears.

13. Validation logic prevents patterns that intersect each other across all the applications.

14. The entire match pattern must validate as a standard URL after adding an http:// prefix and replacing all wildcards with arbitrary words. For example, if the user enters the pattern *.example.com/foo/* the expanded form http://www.example.com/foo/bar can be easily validated using a standard URL validation library.

15. Entering only a domain name for a host part like example.com is valid. It will not match, however, the URL name with a hostname like www.example.com.

16. The form logic auto-corrects the pattern entered by the user, it does this immediately after the textbox loses focus, and the user interface allows the user to see the change before the form is submitted.

17. A pattern intersects another pattern if at least one URL that is matched by both patterns exists. Explicitly, patterns A1 and B1 intersect if all of the following conditions are true:

    1. After trimming the eventual wildcard from the beginning of the host parts, A1.host ends with B1.host or B1.host ends with A1.host.

    2. After trimming the wildcard from the end of the path, A1.path starts with B1.path or B1.path starts with A1.path.

Examples of valid patterns

1. www.example.com/* matches all URLs that start with http://www.example.com/ or https://www.example.com/.

2. *.example.com/* matches example.com, www.example.com, intra.example.com, support.intra.example.com, …

3. *.intra.example.com/* matches intra.example.com, support.intra.example.com, ...

4. example.com/* matches only *://example.com/*.

5. www.example.com/foo/* matches all URLs that start with *://www.example.com/foo/.

6. myintranet/* is a valid pattern for an internal application.

7. www.example.com/foo/* is an existing app match pattern. In this case, the following patterns can be used for any of the applications:

   1. *.example.com/bar/*

   2. www.example.com/foobar/*

   3. *.another-example.com/foo/*

   4. test.example.com/foo/*

Examples of invalid patterns

1. www.example.*/*

2. https://*.example.com/*  because of the scheme

3. www.*.com/*

4. www.*.example.com/*

5. www.example.com/foo?bar because query string components are not supported

6. */*

7. www.example.com/foo#bar because hash components are not supported in the application URL

8. *.example:1234.com/* because of the port (use *.example.com/* instead, all ports are supported)

9. *.example.com/#foo

10. *example.com/* because there is no dot following the star in the host component

11. www.example.com/foo/*/bar/* only one wildcard is allowed in the path

12. *.customer.*.example.com/* only one wildcard is allowed in the host part

13. www.example.com/foo* is an existing app match pattern. In this case, you cannot use the following intersecting patterns for any of the applications:

    1. *.example.com/foo/*

    2. www.example.com/foobar/*

    3. www.example.com/fo*

    4. www.example.com/foo/bar/*

Examples of auto-corrected patterns

1. www.example.com changes to *://www.example.com/*

2. www.example.com/ changes to *://www.example.com/*

3. www.example.com/foo/ changes to *://www.example.com/foo/*

4. www.example.com/foo changes to *://www.example.com/foo/*

5. example.com changes to *://example.com/*

6. *.example.com changes to *://*.example.com/*

A specific use case for non-standard ports

Access the corporate document management system using the following URL:

https://prod.doxydoc.local:32890/start.php

During the business application configuration, you must use the following pattern:

*://prod.doxydoc.local/start.php*

Collecting URLs

URL table is one of the widgets displayed on the dashboard for Speed and Reliability. By default, the system does not store URLs, as they may contain private or sensitive information, especially in custom-written applications. If this is not a concern, Nexthink recommends activating the Collect URLs checkbox as it helps with Key pages configuration. Refer to the Key pages documentation for more information.

• When you enable Collect URLs, users with appropriate permissions can view the individual URLs that employees visit.

• When you disable it, the system does not collect any URLs at the extension level.

• The system renders a truncated URL to preserve data privacy using a sanitizer.

Sanitizing URLs

Whenever possible, Nexthink sanitizes all collected Uniform Resource Locators (URLs) to conceal sensitive information. This is done in an attempt to remove personal data, potential configuration secrets and certain query string patterns.

A URL is a string that serves as a unique identifier for a web resource. A URL consists of four primary components:

• The origin includes the protocol ( http or https), the hostname ( http://nexthink.com ) and the port number (80 or 443), for example, https://nexthink.com:443.

• The pathname specifies the exact location of the resource on the server, for example /company/about-us.

• The hash is an optional component of the URL that stores information for the browser, such as the current state of the web page or user preferences. It also begins with a hash (#) and is followed by a string, for example, in the URL confluence.nexthink.com/pages/urlsanitisaation#url-sanitisation-rules, the hash segment is #url-sanitisation-rules.

• The query parameters are optional pairs of keys and values that provide additional information to the server. They start with a question mark (?) and are separated by ampersands (&), for example, ?projectKey=DCBSM&view=detail. The query parameters are systematically sanitized. The system replaces each key-value pair with [key], for example:

  • https://confluence.intra.nexthink.com/pages/viewrecentblogposts.action?key=AppExEP sanitized with [key]:
    https://confluence.intra.nexthink.com/pages/viewrecentblogposts.action?[key]

URL sanitization mechanism

The URL sanitization mechanism is intended to strip sensitive information from URLs before transmitting them to the server by:

1. Initially removing all query parameters, such as ?key1=value1&key2=value2, retaining only the keys enclosed in square brackets, like ?[key1]&[key2].

2. Replacing any other anchors in the hash segment with the code [anchor].

3. Undergoing further URL sanitization based on specific patterns as described below.

4. Finally, truncating any excessive parts of the URL.

Remember that this process ensures that sensitive data is not inadvertently exposed in URLs.

Pattern matches and tag replacements

Examples of general codes the system uses to sanitize an application’s URL:

These patterns are specified globally for all applications. If any part of the Sanitized URL matches any of the patterns, it will be replaced by the corresponding tag.

Examples:

• https://community.nexthink.com/s/profile/0052p00000BURowAAH sanitized with [hex]:
  https://community.nexthink.com/s/profile/[hex]

• https://outlook.office.com/mail/inbox/id/AAQkAGJhYmFkMTUxLTQ5NzgtNDNlZi1iZDkzLTQ2YzEwNDIwYzA0YgAQAKafeoP2TMVDozSlcBq2JGU%3D sanitized with [gib]:
  https://outlook.office.com/mail/inbox/id/[gib]

URL truncation

There is a maximum acceptable length for the Sanitized URL. If the Sanitized URL exceeds the limit, Nexthink removes the excess and adds an ellipsis (...) to show that some of the URL is missing.

Summary

The following table shows the process of sanitizing URLs:

Activating Soft navigations

Activating soft navigations enables the measurement of the speed of asynchronous page loads where the browser does not load a new page. They are very common in single-page applications (SPA). Soft navigations measure the time it takes a page to stabilize. Refer to the Speed documentation for more information.

Some web applications work in such a way that the page is never completely stable due to background processes. In such cases, soft navigation measurements can report longer than usual page load times. This is why soft navigations are disabled by default. Turn the feature on only when you verify that the application page is stabilized.

Connecting with existing campaigns

Nexthink Applications makes it possible to connect with existing campaigns.

To make the connection, pick the unique identifier (UID) from Campaigns and paste it into the Engage campaign UID text field of the application configuration screen. Note that an Engage campaign license is required.

You can also access the UID value by choosing Campaigns from the main menu and selecting the campaign you wish to connect to from the list. You can locate the UID of the campaign as part of the URL.

Once the campaign is linked, the Sentiment tab on the Applications page becomes visible. Note that you must have the right permissions to access it.

Clicking on the Sentiment tab opens the relevant Campaigns dashboard in a new tab.

RELATED TASKS

• Key pages

• Transactions

• Application licenses

• Thresholds