(Optional, string) Comma-separated list of data streams, indices, and aliases to search. Supports wildcards (*). To search all data streams or indices, omit this parameter or use * or _all.
Failure may be true for example for updating a single doc by id, but an _update_by_query succeeds when run on an index alias.
So am I missing anything or is the Alias documentation not accurate?
It is correct as update by query first runs a query to identify which documents are to be updated (and indices these are located in) before updating these directly using the appropriate index, not the alias.
I see, but it's a bit misleading, since as far as the end user, both of these are considered write operations using an alias: POST alias/_update/5 POST alias/_update_by_query
The way that Elasticsearch handles it behind the scene may be different as you say, but the Alias docs saying that using the alias for write operations will fail as a whole is misleading, and I think that part needs to be revised.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.