Versions Compared

Version Old Version 3 New Version Current
Changes made by

Magnus Frank

Magnus Frank

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

In order for Solr to work, the following parameters are needed:

The goes into the Digizuite web.config in the Appsettings section.

...

The first set of configuration parameters are found in the Digizuite configuration area (System tools/Digizuite Configuration. These are two constants that may be changed:

Name

Description

Default value

SEARCHPROXY_DEBOUNCE_STRATEGY

This parameters is used to determine which debounce strategy to use when updating the index.

The options are:

  • DEBOUNCE

  • THROTTLING

The first one debounces the events that update the index for X seconds and if another event arrives 
for the specific ID, then the debouncer interval X is restarted. This means the index is only updated
once when the system settles for the specific ID.

The latter one is a throttling strategy, where all events are delayed for X seconds. 

Info

The recommendation is to use the DEBOUNCE strategy

DEBOUNCE

SEARCHPROXY_DEBOUNCE_TIME

This parameter is used to determine the time an event is debounced/throttled. 

This specifies the X seconds mentioned the above describtion for SEARCHPROXY_DEBOUNCE_STRATEGY.

The recommendation here varies on the environment. There are the following recommendations:

  • If the environment is in a migration phase (i.e. a lot of data is updated), this should have a higher value

  • The better the environment, the lower a value you want.

The idea is to keep this is low as possible, without spamming the system. If a lot of data is being updated 
on the same asset, many events are spawned. Each event potentially updates the index. The idea is to 
debounce the events just enough, to collect all updates into one index update.

5


Lastly, we have the option to change the behavior of how Solr interprets indexes and queries. This is done using the concept of Tokenizers

The idea of a Tokenizers is to break

...

a stream of characters into a set of tokens. There are many reasons to use tokenizers instead of standard string matching for example performance and/or behavior of search.

For a more detailed explanation of Tokenizers we refer to Solr’s documentation. Additionally for a full list of available tokenizers, we refer to this page

Tokenizers are configured in the following way:

  1. Stop SolrServices (SolrJetty/SolrApache) on webserver

  2. Find the Solr core you want to change (Default location: C:\Bitnami\solr-8.3.1-1\apache-solr\server\solr)

  3. Open <CoreName>\Conf\managed-schema

  4. Find the input parameter you want to manipulate

  5. Change either the type of the input parameter or manipulate the type definition. Both are find the managed-schema

  6. Delete the <CoreName>\Data folder

  7. Start SolrServices

  8. Repopulate the search from Digizuite UI.