Content Delivery search feature scaling
The Content Delivery search feature can be scaled out in various ways, depending on your type of environment.
- All-in-one setup
-
On Development and Test environments, it makes sense to combine your indexing and querying subfeatures in the single IQ Combined Index and Query Services microservice, and to install both alongside all your other microservices on the same machine.
- Dedicated machine for IQ Query Service
-
Split off only the IQ Query Service to a dedicated machine if you work in an environment where your presentation website has little traffic overall, and relatively many visitors compared to relatively few updates.
- Dedicated IQ Combined Index and Query Service or dedicated IQ Index Service
-
For a medium-load Cloud-based presentation website, one dedicated machine for both indexing and querying can prove to be sufficient and cost-effective.
- Full scale-out
-
Giving each IQ service its own dedicated server is recommended for actual high-load production environments, whether hosting on-premise or in the Cloud.
- Multi-instance full scale-out
-
Depending on your experience with indexing and querying performance on a fully scaled-out system, you may need to conduct a cost-benefit analysis to see if setting up a load-balanced system is worth your while. A load-balanced system might involve any or all of the following:
- multiple IQ Query Service instances behind a load balancer
- multiple IQ Index Service instances behind a load balancer
- multiple Elasticsearch instances behind a load balancer