Indexing the whole site to SOLR | Community
Skip to main content
Level 2
Sep 27, 2021 1:58 PM UTC
Solved

Indexing the whole site to SOLR

Currently we are trying to index our site (with 40K pages) using a scheduler Job to SOLR server.

I have tried web page scraping using HTMLUnit and Jsoup, but both approaches take 10+s to form the required model object to be sent to SOLR.

I was able to form the model object using ModelExporter (getting jcr:content as JSON) within 1s. This works fine for single page. But when I run using scheduler (which iterates over the pages), it takes 2-3s. 

so the full site indexing takes 24 hours.

 

Does anyone has any idea on how to do this optimally or any AEM server activity which can speed this up ?

This post is no longer active and is closed to new replies. Need help? Start a new post to ask your question.
joerghoh
joerghohAccepted solution
Employee
Sep 27, 2021 6:11 PM UTC

I don't think that there is a faster way to extract this information in a structured way, but you can always run this process in a multi-threaded way. And instead of just thinking of the initial filling of the index, please consider the cases of updates during regular operation.

3 replies

joerghoh
joerghohAccepted solution
Employee
Sep 27, 2021 6:11 PM UTC

I don't think that there is a faster way to extract this information in a structured way, but you can always run this process in a multi-threaded way. And instead of just thinking of the initial filling of the index, please consider the cases of updates during regular operation.

Translate
Level 2
Sep 27, 2021 2:34 PM UTC

Hi @kiran_vedantam , our old approach (before using model exporter) is from the above links. This took 8s for get the page data. hence moved to model exporter.

Translate