Skip to main content
Version: Omada Identity on-premises 16.0.1

Data object search

You can search in the data of data objects in two ways:

  • Global search – available in the header of the site.
  • Column search – available in list views, where supported.

The search data for a data object is updated every time the data object is saved, and when the display names for its data object type are updated.

info

Auto-complete in a reference property field searches differently from column search and global search. First, it searches the display names of the referenced objects, using right truncation, and returns any matches. Only if no display name matches does it perform a global search on all properties of the referenced objects.

Indexing of search data​

In addition to the sequence number and display name of the data object, the following data is indexed as search data:

Property typeWhat is indexed (data type)
VALUE
  • Text
  • Integer
  • Hyperlink
  • MultiLangText – Values for enabled languages only
  • Log text (if the data type is text, and the property has a log defined)
SET
  • Selected values (Values for all languages)
REFERENCE
  • Display names of selected values
  • Attribute values – if the reference property contains any

For the following data object types, additional data is indexed:

Data object typeWhat is indexed
User
  • User name
Process
  • Sequence number
  • Display name of the target object
Activity
  • Name and display name of the activity assignees
  • Display name of the target object

The following data is not indexed:

  • Properties of the data type DateTime, Decimal, Boolean, or XML.
  • Data object type property bindings that have security defined.
  • Data object type property bindings that are excluded from search.

You can manually exclude a specific data object type property binding from the search index by selecting the Exclude from search checkbox in the data object type property binding.

warning

All properties are included in the search index by default. Exclude properties that are not relevant for search. This keeps the search data smaller and search results more relevant.

Excluding a property doesn't remove its values from search right away. Existing data objects keep the old search data until it is regenerated, for example, when the data object is saved again.

Search modes​

All search modes use right truncation. This means that a search term matches the beginning of a word, not text in the middle of it. For example, searching for Park finds Parker, but searching for ark does not.

Standard SQL search on-prem​

Searches are performed using the SQL LIKE operator. In large databases, the standard search mode can be significantly slower than full-text search. We don't recommend using this search mode in a production environment.

This is the default search mode. It uses the full-text search index of the searchdata column in the tbldataobject table. This mode requires that:

  • The customer setting DisableFTISearch is set to false.
  • The full-text search indexes have been created in the database. This is done automatically, unless the Full-Text Search engine isn't installed in SQL Server.
note

Only Platform administrators can edit the DisableFTISearch setting.

Search results depend on two separate steps:

  1. Generating search data – every time a data object is saved, Omada Identity builds a search string for it from the properties listed in Indexing of search data. Omada Identity performs this step itself, independently of the search mode used afterward.
  2. Executing the search – the generated search string is then matched against your search term using one of the search modes described above.

Because tokenization can happen in either step, some special characters affect search results in ways that are not obvious from the search term alone. For example, an underscore (_) in a value can split it into separate search tokens, so a value such as Project_X may not always be matched by searching for the full, unbroken term.

note

Omada does not currently publish a complete list of special characters that affect search tokenization. If your organization relies on special characters, symbols, or Unicode characters (including emoji) in data object properties, test search behavior for those values in your own environment.

Unicode symbols and emoji in Display name​

Some customers include Unicode symbols or emoji (for example, decorative icons) in a data object's Display name. In some cases, a data object with such a display name cannot be found by searching for the surrounding plain text alone. The symbol or emoji must be included in the search term for the object to be found. For details, see Global search and display names containing Unicode symbols or emoji.

Recommendation: if a data object needs to remain findable by plain-text search, avoid starting its display name with a Unicode symbol or emoji. Because search uses right truncation (matching only from the beginning of a word), a leading symbol or emoji can prevent the rest of the name from being matched by a plain-text search. Where possible, place decorative symbols or icons at the end of the display name, or keep them in a separate property that isn't relied on for search.

Implicit and operator

In the global search, an and operator is placed between all words separated by a space. For example, searching for William Parker results in a search for William and Parker.

or operator

You can search for multiple words with an or operator between them. For example, searching for William or Parker finds data objects that match either word.

The or operator is case-insensitive, and it applies only to the words directly before and after it. For example, searching for William Parker or Kenneth Johnson results in the following search: William and (Parker or Kenneth) and Johnson.

not operator (-)

A - character in front of a search word excludes that word from the search. For example, searching for William-Parker results in the following search: William and not Parker.

Search using quotes

You can enclose a phrase in quotes to search for that phrase in the indexed data.

Operator rules

The following rules apply to the use of operators:

  • You can't use the or and not operators together.
  • You can't use the not operator in front of a quoted phrase.
  • You can't use the or operator in front of the first search word.

Quick filter​

The quick filter is a text field that filters the values in several columns at once. You can use it to search for one or more values.

Quick filter text field with the placeholder 'Search for one or more values', next to the Columns, Filters, Density, and Export options.

When you start entering text, the quick filter returns rows where any of the columns contain a value that starts with that text. For example, if you type the letter "g", you get results from various columns starting with this letter.

You can find the quick filter, for example, in the Access request process when you select a resource and add attributes. A side pane opens with the following grid:

Request access page with the Resources side pane open. The quick filter is highlighted above a grid of resources with Resource ID, Name, Resource type, System, and Description columns.
note

The quick filter is also available in all list views and in side panes for forms, for example:

  • Countries

    Countries list view with 'g' entered in the quick filter, showing Germany (DEU) and United Kingdom (GBR) as results.
  • Tasks

    Tasks list view with the quick filter highlighted above the Name, Process, Target, and Active time columns.
  • Identity owners

    Identity owner side pane with 'Anne' entered in the quick filter, showing matching identities such as Anne Henning, Anne Stewart, and Annette Sturm.