Re: Design question about Collection operations

2024-12-03 Thread David Smiley
BTW when I speak of "core level", it's core/configSet since the configuration applies to all cores of that type and they literally run in the context of a core. On Tue, Dec 3, 2024 at 11:53 PM David Smiley wrote: > No encryption isn't different from one core to the next but when you tell > a col

Re: Design question about Collection operations

2024-12-03 Thread David Smiley
No encryption isn't different from one core to the next but when you tell a collection to encrypt itself, you expect all cores to participate in this. I don't think of this as being a factor in choosing solrconfig.xml registration vs Collection API. At least core level is easily pluggable in Solr

Re: Design question about Collection operations

2024-12-03 Thread Jason Gerlowski
> There is ambiguity of the idea of a core or even collection "admin" > handler. Ah, I see. I've always used the terms "core admin" and "collection admin" to refer to the specific endpoints exposed at those paths (i.e. /admin/cores and /admin/collections). But you're right that it's ambiguous - I

Re: Design question about Collection operations

2024-12-03 Thread David Smiley
When Bruno said a "Core level handler", he meant a RequestHandler registered in solrconfig.xml and thus accessible as /solr/coreName/admin/encrypt as seen here: https://github.com/apache/solr-sandbox/blob/4e1819ff8a3758becca19bf337ecd1b352dba805/encryption/src/test/resources/configs/collection1/sol

Re: Design question about Collection operations

2024-12-03 Thread Jason Gerlowski
I think the most common way we do stuff like this is two have an API at both the "core" and "collection" levels, with the "collection" API being used to orchestrate N calls to the lower-level "core" API. Using encryption as the example: the "core"-level API might manage the encryption of a single