murblanc commented on pull request #1694: URL: https://github.com/apache/lucene-solr/pull/1694#issuecomment-664312074
Thanks Noble. So what’s your guidance regarding the Autoscaling PR? Shall we assume the Autoscaling plugins need to use these interfaces or shall I continue building cluster/node/collection/shard/replica abstractions that make plugin development easier? If the first option, there will be a few requirements on these interfaces (and their implementations), for example allowing them to be keys in maps (variable values or snitches are per node or collection or replica etc, they need to be stored somewhere so the plugin can access them). Ideally also as stated elsewhere, all these interfaces implement a common one (could be empty) so that snitch/variable target can be passed easily. The Autoscaling plugin work can also define similar abstractions on top of the interfaces defined here. I do see value for all plugins of all types to share a single set of abstractions but obviously this puts more constraints on these abstractions... ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
