L'introduction de ES|QL dans Elasticsearch facilite la recherche et l'analyse de grands jeux de données.\n\nES|QL présente ses résultats sous forme tabulaire en JSON, CSV et aussi au format Apache Arrow, un format de dataframe compact permettant des échanges sans désérialisation, qui est nativement supporté par la librairie Python Pandas.\n\nCette intégration ouvre de nouvelles perspectives pour l'exploration des données avec les outils habituels des data analysts, et l'intégration facile des pipelines d'aggrégation dans les applications.\n\nAprès un bref aperçu de ES|QL, nous ferons une exploration interactive d'un jeu de données avec ES|QL, Arrow et Pandas dans un notebook Jupyter. Et un petit benchmark vous montrera l'efficacité du format Arrow comparé à JSON !

ES|QL is a new piped query language for Elasticsearch. It supports writing composable queries and it features a multi-staged execution. Unlike the other languages supported by Elasticsearch, ES|QL doesn't transpile to Query DSL or use the internal search client: it's based on its own stack. This comes with a sophisticated query analysis and optimisation steps, as well as parallelisation and vectorisation. This talk will give an overview of the execution flow of a query and touch on a few key implementation aspects, following the query from its first syntactic analysis down to Lucene delegation followed by returning the results back to the user, all in a distributed environment.

Elasticsearch and Kibana added a new query language: ES|QL — coming with a new endpoint (_query) and a brand new syntax. It lets you refine your results one step at a time and adds new features like data enrichment and processing right in your query. And you can use it across the Elastic Stack — from the Elasticsearch API to Discover and Alerting in Kibana. But the biggest change is behind the scenes: using a new compute engine that was built with performance in mind. Join us for a quick overview and a look at syntax and internals.