PT 8.62

Randall Groncki

Scheduling Search Indexes moves our data from PeopleSoft to OpenSearch.  Once OpenSearch has ingested that data, we can then use it in our Visualizations.

Prerequisites:

1. PSQuery to collect and shape our data
2. Deployed Search Definition that controls that data

This post follows these previous posts and uses the objects we created.   This post explicitly refers to those objects to continuing the lesson.  Review these posts if needed before working through this post.

Extract Data

Navigation: Home > PeopleTools > Search Framework > Schedule Search Index

Add a new Run Control ID.

We will reuse these run controls often, so we’ll name it something intuitive.   I use the Search Definition name and a suffix of “_FULL” or “_INC” depending on the extraction type.

For this example, I’m using the Search Definition “X_PT3_TRAINING” with the suffix “_FULL” since this will be a full index update.

X_PT3_TRAINING_FULL

Search Instance:

The Search instance should default to the system defined Search Instance.

Choose the Search Index we created in the last post: X_PT3_TRAINING

Search Definition:

Choose the Search Definition

X_PT3_TRAINING

Language Options Box

Leave the default value of “Base Language”

Index Type

Change the index type to “Full Index”

A “Full Index” is the easiest to work with.  Any data in this current data set in OpenSearch will be deleted and the data from this extract will replace it.

An “Incremental” index uses the prompt on the Last Update Datetime field to pull only the data rows with a later datetime value than the last time this extract was run for this Search Definition.  PeopleSoft tracks when each search definition has run.  

During an incremental update, no documents in OpenSearch are deleted.  Only new documents from this extract are added.   If a new document has the same Drilling URL as an existing document, then the new document replaces the older document. 

A Full Index is simpler and easier to manage.   Depending on the size of your data, it could take longer to execute.  This could also interrupt you user’s experience if they are using a dashboard as you are running a full index.

Run

Run the process in the process scheduler and wait for the results.

Here is the Message Log from our process.   We can see in the log results that 237 documents were sent to OpenSearch. Reference your Query to ensure that this is the same number from your Query running for all data.

Wrap Up

If your process completed successfully and you see your “Total Documents Indexed” message, your data should now be in OpenSearch.

Next

In our next blog and video, we’ll cross over into Insights and create a Search Definition for this data that is now in OpenSearch.