How We Have Improved Searching Within Ticketsolve

Elasticsearch, will improve the searchability of shows for both patrons and back-end users of Ticketsolve.

We want to take a minute to explain what it is how it is going to improve searches in the system.

First, we need to take a step back and explain a little about data storage and ticketing. Data can be stored in several forms. How it gets stored depends on the type of work the user need to be perform. So for example, there may be data that you store that needs to be accessed frequently, data that needs to be accessed as a once off and data that needs to be accessed once in a while. In addition to this, some data needs to be search very precisely and specifically, while other data may be searched very generally, to have a broad view. So there is a lot at play when it comes to storing and accessing the data required, and how search terms are used.

Previously, Ticketsolve used MySQL for storing and searching the show data. While MySQL has many advanced functionalities, we felt we needed to improve textual searches. We have chosen Elasticsearch, which is a flexible and powerful open sourced, distributed, real-time search and analytics engine for the cloud.

Elasticsearch will give the Ticketsolve community and their patrons several improvements:

• Faster searches
• More flexible searches: For example, a search may be adjusted to use a generic term, say ‘book’, to return a broad result, like ‘book’, ‘books’, ‘bookkeeper’ and so on.
• Leaner and faster internal operations: extending the data structure, for example to improve a certain search, is going to be faster and easier.
• Search results will now show the matching title first, rather than the generic ordering that was done previously.
• Extending Elasticsearch to also give a higher score to, say, tags, is going to be a simpler operation now.

We have deployed Elasticsearch for shows at the moment.

We would love to hear if you have have seen any improvements with the new search for shows. Let us know!