---json
{
    "canonical": ":analytics:replication",
    "description": "The analysis and replication of a webpage are the cornerstones of the quality and performance of a page. This page explains how this is done.",
    "name": "Database Replication",
    "page_id": "670cgbvcwnweskpj0g3j5",
    "title": "How ComboStrap analyzes and replicates data ?"
}
---
====== ComboStrap - Analytics and database replication ======


===== About =====
The ''database replication process'' is the process that:
  * creates the [[data|analytical data]] if needed
  * replicates the data into the [[docs:analytics:database|database]].



===== Sync =====
Because this is an intensive process, it happens slowly:
  * for every page visited
  * or for every page that gains or lost backlinks.
And the output is [[:docs:cache:cache|cached]].


If you want to sync the database completely, you can:
  * run the [[#cli|cli]]
  * run the [[#search index manager|Search Index Manager]]



==== Search Index Manager ====

The ''search index manager'' is a plugin that will replicate the page data. See how to install it and run it on this page. [[:docs:utility:search_index_ manager]]

==== Cli ====
On the server, you can [[docs:admin:replication|replicate the data]] to the database with the [[:docs:admin:cli|cli]] following command
<code bash>
cd $DOKUWIKI_HOME
# optional (for farm)
animal=animal-directory-name
# command
php ./bin/plugin.php combo metadata-to-database --host serverHostName --port 80 /
</code>

===== Monitoring =====

==== The replication date ====

The replication date is stored in the ''date_replication'' metadata and can be seen via:
  * the [[docs:metadata:manager|metadata manager]] (integration tab)
  * the [[docs:metadata:viewer|Metadata Viewer]].
  * or [[query|query]] .
<code sql>
select
	path,
	date_replication
from
	pages
order by date desc
limit 10;
</code>


==== The analytics date ====

You can see the analytics date on the ''date'' field of the [[data|JSON object]]

Example of [[query|query]] .
<code sql>
select
	path,
	json_extract(analytics, '$.date') as date
from
	pages
order by date desc
limit 10;
</code>

==== Replication Request ====
When [[docs:navigation:backlink|backlinks]] are added or deleted, a replication request of the page is asked to update them.


This request are stored in the table ''pages_to_replicate''. If you query this table, you would get all page that should be replicated in the near future. This table should be empty 99% of the time as this is a temporary table.
<code sql>
select * from pages_to_replicate
</code>