Introduction

Indexes inside BravoSearch indicate the lowest common bucket for the solution. The index is where all of your data is stored and shows the way in which it is structured for storage. An index stores types that are the best representation of your data as it can be processed by BravoSearch. Information in the indexes are bound to types that represent the data streams that are being ingested by BravoSearch, the fields represent the ingestible data in that type.

Index Fields

[back to top]

The First thing that you will see when you come to the indexes page is a list of all of the indexes that are currently associated to the account that you are associated with. Additionally there will be a list of the number of documents that are currently stored in that index. 

The index fields are a list of all of the types and fields that are currently associated with the index.  This will allow you to see what Types are being logged in the system as well as the fields that are associated with those types.  Types generally are aligned with the collectors, so there may be a collector for HigherLogic Discussions in which case you will see the type "HigherLogicDiscussions" listed in the index information.

The column for Indexed indicates if the information is stored in the system for searching.  In most cases this will always be true.

The column for Stored indicated if the information is stored in the system for use, for static data to be indexed it must be stored.  In most cases this will be true. Instances where we have calculated fields may indicate an instance where something is not stored.

The column for Analyzer indicates any analyzers that have been applied to the field, these are all controlled by convention and can be explained as follows:

  • Dynamic -Standard search analyzer using all of the command's default analyzer and token settings. This includes synonym graph filtering.
  • Unanalysed String - Unanalyzed string, indexing the entire field value as a single token. This is most useful for term aggregations (facets).
  • Unanalysed Integer - Unanalyzed field forcing all input to be treated as an int data type.
  • Unanalysed Date - Unanalyzed field forcing all input to be treated as date data type. Supports ISO 8601 extended and unix timestamp format (millis).
  • Keyword Autocomplete - Unanalyzed field enabling completion functionality. This field is unanalyzed, completion will happen on the entire field value as a single token.
  • [Obsolete] Searched Autocomplete - Unanalyzed field enabling completion functionality. This was a feature of a previous search version. Existing instances will default to the Keyword Autocomplete behavior

Index Utilities

[back to top]

The index utilities menu is included as a set of tool for interacting with the index directly, these actions should only be performed by individuals understanding the full consequences of each of the actions.

Truncate Index is included as a last ditch effort in the event that something has gone wrong with your instance of an index.  It will make a call to bravo command to clear all of the items out of the index and start collecting again from the beginning. WARNING: when you press this button there will be no contents in your index until the next collection,  so when someone uses your search there will be zero results until the next collection.  This can be rectified by going to the collection screen and manually kicking off collection, however for the delta between when the truncation is executed and the collection completes there will be zero, or a partial number of records.  This is not a mistake, this is the nature of the way that search works.