Towards an open source tool stack for e-commerce search
Chorus makes deploying powerful ecommerce search easier by shifting the buy vs build decision in favour of build, so you can own your search! Chorus deals with three issues:
-
Starting from Scratch is Time Consuming Downloading an open source search engine isn't enough. It's like beeing provided with all the parts of a Lego model, but without building instructions ;-) . We need a better baseline to quickly get started.
-
Integration of Tooling is Hard Search isn't just a query running on an index of data. It's also the operational monitoring, the analytics and relevance tooling, that goes into it. Every team that starts, faces the penalties of starting from scratch, integrating an ecosystem of options.
-
Sharing Knowledge is a Must! It isn't enough to just have conference talks. We need code and data samples in order to share knowledge about improving ecommerce search. Chorus is a public environment that you can use to profit from the community and also to share your next great idea!
This is the project that enables Chorus to use Elasticsearch as the search engine. As of now (March 2022), this project is under development. For those interested in the whole stack: You can visit the Solr version of Chorus
Want to stay up-to-date with the community? Visit https://querqy.org/ to learn more, and join the E-Commerce Search Slack group for tips, tricks, help and news on what's new in the Chorus ecosystem.
- 23 March 2022: Chorus, now also for Elasticsearch!
- 17th June 2021: Encores? - Going beyond matching and ranking of search results - Chorus is used at BerlinBuzzwords.
- 15th November 2020: Chorus Workshop Series Announced - Learn from the creators of the components of Chorus via six workshops.
- 17th October 2020: Chorus featured at ApacheCon @Home - René and Eric give a talk at ApacheCon on Chorus.
- 10th June 2020: Chorus Announced at BerlinBuzzwords - First release of Chorus shared with the world at a workshop.
- April 2020: Paul Maria Bartusch, René Kriegler, Johannes Peter & Eric Pugh brainstorm challenges with search teams adopting technologies like Querqy and come up with the Chorus idea.
- The UI (Reactivesearch) runs at http://localhost:4000 | http://chorus-es-edition.dev.o19s.com:4000
- Elasticsearch runs at http://localhost:9200 | http://chorus-es-edition.dev.o19s.com:9200
- Kibana runs at http://localhost:5601 | http://chorus-es-edition.dev.o19s.com:5601
- SMUI runs at http://localhost:9000 | http://chorus-es-edition.dev.o19s.com:9000
- Quepid runs at http://localhost:3000 | http://chorus-es-edition.dev.o19s.com:3000
- Keycloak runs at http://keycloak:9080 | http://chorus-es-edition.dev.o19s.com:9080
- Prometheus runs at http://localhost:9090 | http://chorus-es-edition.dev.o19s.com:9090
- Grafana runs at http://localhost:9091 | http://chorus-es-edition.dev.o19s.com:9091
Working with macOS? Pop open all the tuning related web pages with one terminal command:
open http://localhost:4000 http://localhost:9200 http://localhost:5601 http://localhost:9000 http://localhost:3000
We are trying to strike a balance between making the setup process as easy and fool proof as possible, with the need to not hide too much of the interactions between the projects that make up Chorus.
If you are impatient, we provide a quick start script, ./quickstart.sh that sets Chorus up for you. However, I recommend you go through Kata 0: Setting up Chorus, to get a picture of what's running in the stack.
To start your environment, but still run each command to set up the integrations manually, run:
docker-compose up --build -d
The quickstart script will launch Elasticsearch, download and index the sample product data for the ecommerce index:
./quickstart.sh
If you want to add in the offline lab environment based on Quepid, then tack on the --with-offline-lab parameter:
./quickstart.sh --with-offline-lab
To include the observability features (Grafana, Prometheus, and Elasticsearch Exporter), run:
./quickstart.sh --with-observability
To see what is happening in the Chorus stack you can tail the logs for all the components via:
docker-compose logs -tf
If you want to see the logs of just one component of the Chorus stack, use:
docker-compose ps # list out the names of the components
docker-compose logs -tf elasticsearch # tail elasticsearch only
To stop all containers, you can run:
./quickstart.sh --stop
To destroy your environment (including any volumes created, like the mysql DBs), just run:
docker-compose down -v
or:
./quickstart.sh --shutdown
If Docker is giving you a hard time, then some options are:
docker system prune # removes orphaned images, networks, etc.
docker system prune -a --volumes # removes all images, clears out your Docker diskspace if you full.
You may also have to increase the resources given to Docker - up to 4 GB RAM and 2 GB Swap space.
The Chorus project includes some public, sample datasets. These datasets enables the community to learn, experiment and collaborate in a safe manner and are a key part of demonstrating, how to build measurable and tunable ecommerce search, with open source components.
The sample product data is generously sourced from Icecat and is licensed under their Open Content License.
The version of the Icecat product data that Chorus provides has the following changes:
- Data converted to JSON format.
- Products that don't have a 500x500 pixel image listed are removed.
- The Prices of ~19,000 products got extracted from the https://www.upcitemdb.com/ service, using EAN codes.
-
Keycloak apparently does not run on Apple M1 chips with the pre-built images. To avoid this issue, don't use Keycloak authentication for Quepid. For a workaround please follow these instructions: docker/for-mac#5310 (comment)
The workaround has been successfully tested with Keycloak version 16.1.1.
-
SMUI stands for 'search management UI'. It is designed to work with Solr. We provided some scripts for basic functionality with Elasticsearch but there are still limitations. You get a feeling what's currently possible by going through Kata 1: Optimize a Query.
-
RRE: The technical integration is able to run queries and get the correct results from Elasticsearch, but apparently it computes the search metrics incorrectly.
Of course, contributions are very welcome to improve Chorus - The Elasticsearch Edition!