System Settings – Search tab

Note: You must configure or create a number of Dynamic System Settings to enable search facilities in Pega 7 Platform. Verify that the settings have the values specified in the table at the end of this topic.

Index administration

Use the Index administration settings to enable or disable indexing for classes under Rules-, Data-, and/or Work-.

Note: Through a check box labeled Exclude this class from search? on the Advanced tab of the Class form, you can cause full-text searching to skip over instances of a specific class. Search results then exclude any instances of that class.

Select Enable indexing to turn on search indexing.
Select All rules to turn on indexing for all Rule-classes.
Select All data to turn on indexing for all Data- classes.
Select All work to turn on indexing for all Work-classes. In a production system, building and maintaining search indexes for work objects is typically more costly in performance impact than building and maintaining search indexes for rules or data instances. Work object search is not available in the Designer Studio.

To the right of each check box, the display lists the index file size for that index type.

For each index type:

Click Check to check the status of the index files. In the form that appears, click Run full diagnostic check to launch a full diagnostic check of rules, data or work. This has the greatest chance of identifying issues, but also takes the longest and uses the most resources.
Click Re-index to rebuild the complete index file for the selected type. For the Work- index files, a form appears where you can select
- All classes
- Specific classes that you list
- All classes except those you list
Below the radio buttons where you make your choice is a section where you can list the classes to include or exclude. You can add as many rows as you need, or remove rows you no longer need. This allows you to select a subset of classes to re-index, saving time and using less server capacity.

When you are ready, click OK to re-index; or click Cancel to close the form without re-indexing.

You normally only need to perform these actions when search index files are corrupted or accidentally deleted.

Progress messages appear at the bottom of the page during installation and during re-indexing, including success message when the process is complete for each index. Additional details and progress messages appear in the system console log.

The system creates the indexes for these classes under the directory specified in the indexing/explicitindexdir Dynamic System Setting, as follows:
- specifiedindexpath/PegaRULESIndex/RuleIndex
- specifiedindexpath/PegaRULESIndex/DataIndex
- specifiedindexpath/PegaRULESIndex/WorkIndex
Re-indexing in nodes that belong to a cluster
If you click the Re-Index button in a node that belongs to a cluster, the system records each re-indexing as an instance of the Log-Cluster-FTSIndex class, including the node ID and pyStatusWork, which is set to "Open-InProgress". As a result, the system disables all the Re-Index buttons for the node and for all other nodes that belong to the cluster. After re-indexing completes, pyStatusWork is set to "Resolved-Completed" in case of completion without errors, or to "Resolved-Error" in case of completion with errors.
Notes:
- The pr_log_cluster table holds instances of the Log-Cluster-FTSIndex class derived from the Log-Cluster- class.
- If a node is shut down while its classes are being re-indexed, pyStatusWork retains the value of "Open-InProgress", and the system disables the Re-Index buttons permanently. To enable the buttons, an administrator can run the pzResolveFTSIndexEntry activity for the node ID to set pyStatusWork to "Resolved-Error".
- You cannot re-index from the command line utility if you have initiated re-indexing in the same node or another node in the same cluster from the user interface, and the re-indexing is still in progress.
Click Cancel Re-index to cancel re-indexing that is in progress.

Select Index attachments to include the contents of attachments in indexing. Enable this capability only if your applications provide full-text searching of work objects. PDF, Microsoft Office, and text attachments are indexed and will be full-text searchable. However, if your application saves work item file attachments in an external enterprise content management system accessed through Connect CMS rules, such file attachments are not included in Pega 7 Platform search results.

The Attachments threshold (MB) display holds the maximum size in megabytes of attachments to be indexed. The default, which you can adjust to accommodate your system's needs, is 5 MB as described in the Dynamic System Settings section above. The log file records the existence of attachments that exceed this threshold and that are not indexed.

Note: You can check or clear any of the boxes above while a Pega 7 Platform server node is operating, but such index processing can affect system performance if the indexes are large. As a best practice, choose a time when the server is less busy to perform these one-time processor-intensive operations.

Note: The Designer Studio search facility also allows search of Application Developer help topics. The Help search capability depends on a static index and is unaffected by the settings you record in this tab. During installation or upgrades, the help index file is automatically extracted from a binary file rule named HelpIndex.pxHelpIndexBinary.zip.

Agent information

This section displays the status of the Pega-RULES agents that incrementally update the search index files as changes are made to Rules-, Data-, and/or Work- instances. For each agent the display provides:

Search Index Host Node Setting

In a multi-node cluster system, the Elasticsearch index files required for search may be hosted on one or more nodes.

To enable Elasticsearch to provide high reliability and uptime for search queries, multiple nodes should be configured on the Search landing page as host nodes for Elasticsearch index files.

Note: It is recommended that at least two, and ideally three, nodes be configured as Elasticsearch host nodes for maximum fault-tolerance.

Caution: Do not define additional host nodes (which will trigger the replication of ES indicates to those nodes) until the ES index files have been completely built on the first node, and the upgrade process from Lucene to ES has completed.

This section displays the current list of host nodes for Elasticsearch index files. This information is stored in the pr_sys_statusnodes table in the Pega 7 Platform database.

Click Add Node below the list to add a node. Remove a node by clicking the X icon to the right of its information.

Note: For a new install, allocate at least 1 GB of free space for the Elasticsearch index files. For an upgrade, allocate three times the amount of space used by the Lucene index files. This extra space is recommended to allow Elasticsearch to perform automatic index management in which it merges data in various segments of the index to optimize performance.

Query Settings

Use the settings in this section to change how Elasticsearch works when you select the option to perform a fuzzy search query in a custom search tool. Fuzzy search queries look for approximate spellings that match the entered search string, and find matches when you misspell a term, or are uncertain of its spelling (for example, technical terms and foreign language terms). With the fuzzy query option, entering a term like “tst" matches terms like "test", "tests", and so on, depending on the parameters that control how a fuzzy search query operates:

For more information about how Elasticsearch implements fuzzy search queries, go to the Elasticsearch website at elastic.co.

Automated Search Alerts

You can set automated spot checks to run periodically on the status of search agents and index files, and to report on events and issues that arise. The system sends out alerts when:

Select Automatically Monitor Files to automatically review the index files at the interval you set in the Frequency of Monitoring field.

To send notifications after re-indexing and performing diagnostic checks, select Send Email Notifications on Completion of Re-Indexing and Full Diagnostic Checks.

To send notification when there has been a modification to search host nodes, select Send notification on modification of search host nodes.

If you have selected either of the previous two check boxes, fields appear where you can provide a list of email addresses to send notifications, and the email account to use for sending the messages. Click the icon to the right of this field to set up a new account.

Messages

When you have updated settings, this section displays messages about the changes to parameters you have made.

Click Submit for the changes you have made to take effect, and to build the indexes.

Dynamic System Setting	Default value:	Description
indexing/distributed/http/port	9300-9400	To enable communication between nodes by Elasticsearch, a TCP port should be opened for each node in a Pega 7 Platform cluster and made available for use by Elasticsearch (for internal node-to-node communications only, and not opened to the outside world). This setting can be used to specify a range of ports for Elasticsearch's use.
indexing/distributed/batch/numworkers	2	This setting can be used to control the number of threads on each node that are processing entries in the search queue during bulk indexing operations such as clicking the Re-Index button on the search landing page.
indexing/distributed/search_timeout_millis	N/A	This setting can be used to set a search query timeout limit in milliseconds.
alerts/search/operationtime_threshold	N/A	This setting can be used to generate alerts for search queries whose response time in milliseconds exceeds the specified threshold.