Look for something among the old blog posts.

You might also be interested in this and this.

Thanks RVR. I have gone through those links. Most of them were created 4-5 years back in the pre Elgg 2.x era.

Many of those optimisations become part of the core like simplecache, HTML CSS tidy options, minification etc.

In those posts there are multiple optimization techniques have been mentioned which we generally do for most web applications.

What I was asking about was the hardware configurations, and specific optimization methods (among those mentioned in the list) used here in community website. The community website performance seems amazing.

What I can find about the server specs is

• 2 cores
• 4GB memory
• no memcache
• local DB
• PHP 7.0
• local SOLR instance for search

You can find all the plugins installed on the community site here

https://github.com/Elgg/www.elgg.org

I think the biggest performance gain comes from moving the search to a dedicated search engine. This will speed up the search and unburden the database from all the heavy search queries.

Thanks Jerome for the input. 

Thanks a lot @jerome, 

Can you provide a link explaining how to move the search to a dedicated search engine ?

Regards,

@hermand,

it depends on how a plugin is configured. So there is no general way to describe it. Hopefully the plugin developer of a search plugin will explain the process ;)

For the community site we use Arck Interactives SOLR plugin https://elgg.org/plugins/1635100

There is also a plugin for Elasticsearch by ColdTrick (me) https://elgg.org/plugins/2922714

Thanks

@jerome, I have one more question.

I want to understand how it works.

Let's imagine that I manage to move the search to a dedicated search engine.

Look at this query.

$namevalue_pairs = [];

$searchString = get_input('searchString');

$namevalue_pairs[] = [

'name' => 'entityname1', 

'value' => '%' . $searchString . '%',

'operand' => 'LIKE',

'case_sensitive' => false,

];

$namevalue_pairs[] = [

'name' => 'entityname2', 

'value' => '%' . $searchString . '%',

'operand' => 'LIKE',

'case_sensitive' => false,

];

$options = [

'types' => 'object',

'subtypes' => \MySubType::SUBTYPE,

'full_view' => false,

'list_type_toggle' => false,

'item_class' => 'market-item-list',

'pagination' => true,

'no_results' => elgg_echo('mysubtype:none:found'),

'metadata_name_value_pairs' => $namevalue_pairs,

'metadata_name_value_pairs_operator' => 'AND',

];


$content = elgg_list_entities($options);

Have a look at http://learn.elgg.org/en/stable/guides/search.html