AI OnAI Off
How to use multiple indices for multisite?
Normal solution is one Find index per Episerver installation, multi site or single site...
You can easily restrict search result per site...indexing will be problematic otherwise...
Jun 08, 2016 19:14
The simplest way when doing unified search is to do nothing ;-) The indexing job is multi site aware and will automatically split your index when having a multi site (one partition for each site and one for common content). Also unified search is multi site aware and will automatically filter the result based on the current site of the user request. In case you use the core Find API you can use the .FilterOnCurrentSite()-extension to filter content based on the current site of the user request.
/Henrik
Jun 09, 2016 7:58
Thank you, Daniel and Henrik!
@Henrik, is it correct that it actually does the FilterOnCurrentSite by default? I saw that you had answered that on another forum post.
(Ps. we are currently using version 11 of Find)
Cheers,
Cathinka
Jun 09, 2016 12:48
When using Unified search it defaults to filterForPublicSearch = true in GetResult(...) and that in turn applies a number of filters (FilterForVisitor(...) (filters content on current language and access user access rights), ExcludeContainerPages(...), ExcludeContentFolders(...), FilterOnCurrentSite(...)). This has always been available in Unified search since it was implemented so with Find 11 you are ready to go!
/Henrik
Jun 10, 2016 7:59
This topic was created over six months ago and has been resolved. If you have a similar question, please create a new topic and refer to this one.
Hi,
I am working on a project where we are going from a single site solution to a mulitsite solution.
For the original site we have a standard Epi Find Index. When extending to a multisite, we want to keep search results separate for each site.
I have therfor thought that the cleanest solution is to have one index per site.
How is the best way to set up this in configuration and with regards to the indexing job? And what is a good way to decide which index is used in the code?
We are currently using Epi Find unified search.
Thank you for any pointers :)