Quite a bit here, I'll keep it concise. - Merging everything into one repo - yes absolutely.
- Sidecar does a lot more than give analytics a path. cassandra-analytics is narrowly scoped, it doesn't make sense to have something that does a lot more than analytics be in a repo called analytics. This would be confusing for users. I am deeply against confusing people. - Fine with a new repo cassandra-ecosystem. I woudn't even be opposed to having a mono-repo with all the drivers too or merging them all in as well (just musing about this, thinking about the future). It's broad. Fine with anything that says "this is where you get all the extra stuff" - I put a lot of work into getting GH CI working for analytics & sidecar. I think it's at a good place now, and there's more we could do, BUT I don't know if there's really anything to gain at the moment by removing Circle. For example if someone pays for CircleCI and they just want to get their stuff built and tested, we might as well allow that. GH does have concurrency limits that I've run into already. TL;DR yes to merge, just not into cassandra-analytics keep circle till it's a pain. Jon On Wed, Jun 3, 2026 at 8:14 AM Josh McKenzie <[email protected]> wrote: > I assume we'd keep published artifacts named the same no matter the repo > layout > > That's my thinking. > > I prefer to merge both into a new repo something like > cassandra-analytics-and-sidecar > > I must say, I don't love that name. :) What about something like: > - repo: cassandra-ecosystem > - JIRA: CASSECO > > That'd leave the door open for us to expand it to include other things in > the ecosystem if we found this helped with development velocity and > workflows. > > it makes more sense to me to merge Sidecar into Analytics than the other > way around. > > I disagree on this one but mostly from a gut intuitive perspective and not > something deeply thought out. In my mental model, cassandra-sidecar is > something of a superset of cassandra-analytics in terms of its > functionality / role / code. If we were to zoom out and weigh the > dependency direction between the two projects, my assumption is that > cassandra-analytics depends more heavily on sidecar than the other way > around. Joey's point as well resonates re: the user perspective of > subsuming a specific-purpose tool into a general-purpose tool, should we go > the "merge one into another" route. > > I have a modest preference to the new merged repo of "cassandra-ecosystem" > because of the potential future paths that unlocks for us, but either seem > acceptable to me and a clear improvement on the status quo today. > > And re: CI: I'm good with dropping Circle if everyone else is. I've been > leaning more on the github validation lately anyway. > > On Wed, Jun 3, 2026, at 10:16 AM, Joseph Lynch wrote: > > I think merging repos makes sense but merging analytics (a > specific-purpose tool) into sidecar (a general-purpose tool) might make > more sense from a user perspective, even if a lot of of the complexity > currently lives in analytics. We briefly considered "tools" or "extra" back > in the day, but felt that might not convey the long-running nature of the > process. > > I think it will be confusing to users and contributors if rolling restart > behaviors or health monitoring are contributed to "cassandra-analytics". > > Also just want to double-check: I assume we'd keep published artifacts > named the same no matter the repo layout (sidecar clients, analytics-core, > etc ...)? > > > -Joey > > On Wed, Jun 3, 2026 at 9:41 AM Štefan Miklošovič <[email protected]> > wrote: > > On Wed, Jun 3, 2026 at 1:02 PM Shailaja Koppu <[email protected]> wrote: > > > > Hi Josh, > > > > Thanks for taking care of this. This merge definitely improves > development velocity. Here are my recommendations for outstanding questions. > > > > 1, I prefer to merge both into a new repo something like > cassandra-analytics-and-sidecar, to ensure people don’t think that one of > them vanished or no longer exists/supported > > I think this is just a matter of communication. As said, the > repository of deprecated Sidecar is not going anywhere. We put a big > fat notice that this repo is no longer maintained and all moved to > analytics. People will not think it has vanished because it will be > still present. > > For me the only reason we go with the third repository is to rename > whole cassandra-analytics to something completely different where we > expect software beyond analytics and sidecar to be added in the future > as well. If we start to add things into cassandra-analytics which will > not have necessarily a lot in common while we still wanted to release > it as one thing, maybe creating something like cassandra-ecosystem or > cassandra-platform or similar justifies a third repository to be > created. > > > 2, I prefer the same for Jira as well, something like > CASS-ANALYTICS-AND_SIDECAR or better name if you can come up with one. And > with in that Jira board, separate sub-tags for Analytics and Sidecar, that > way if someone needs only Sidecar tasks, they can apply that sub-tag. And > admins/developers will be able to filter Jiras based on sub-tags. > > 3, Perhaps we can ask authors to migrate their open PRs to the new repo > > 5, Existing repos can be marked as archived and update Readme pages to > redirect incoming developers to the new repo > > > > Thanks, > > Shailaja > > > > > > > On Jun 3, 2026, at 9:30 AM, Štefan Miklošovič <[email protected]> > wrote: > > > > > > More clarification behind that: > > > > > > If we go to merge, it makes more sense to me to merge Sidecar into > > > Analytics than the other way around. Yes, Sidecar also exposes > > > functionality which has nothing to do with Analytics as such but it is > > > still better than merging Analytics into Sidecar, imho. We basically > > > can't use Analytics without Sidecar. Sidecar is inherently part of > > > Analytics. > > > > > > If we were to create a third repository where both are merged, we > > > would need to likely come up with a new name for that project, create > > > a new JIRA subproject ... We would then deprecate both repositories > > > instead of just one and create basically a brand new project. I > > > consider this to be just unnecessary. We would just make everything > > > more complex. > > > > > > From ASF perspective I do not think that we would need to move a > > > repository to deprecate to "Attic". My understanding is that Attic is > > > something where whole projects are moved. We do not have a separate > > > PMC for Sidecar / Analytics so I think that we can just merge these as > > > we want. > > > > > > (1) https://attic.apache.org/ > > > > > > On Wed, Jun 3, 2026 at 10:04 AM Štefan Miklošovič > > > <[email protected]> wrote: > > >> > > >> My (personal) take on this: > > >> > > >> 1) Merging into one repository, no into a new repository. > > >> 2) Retiring CASSSIDECAR and using only CASSANALYTICS. Existing tickets > > >> under CASSSIDECAR can be delivered into a new repository. > > >> 3) Nothing. There are 24 PRs in Sidecar repo, 14 are some "chore". > > >> Converting the rest, even manually, to analytics repo is easy. > > >> 4) maybe we can start to use pre-ci and ci? Why are we so fixated on > Circle? > > >> 5) only cassandra-sidecar will be deprecated, will be marked as read > > >> only / archived, we will not remove it > > >> 6) yes > > >> > > >> On Tue, Jun 2, 2026 at 10:20 PM Josh McKenzie <[email protected]> > wrote: > > >>> > > >>> I'd like to propose we merge the cassandra-sidecar and > cassandra-analytics repositories. I've shopped the idea around to some of > you and gotten universally positive feedback with some questions about > details we deferred to this discussion. > > >>> > > >>> Reasons we should merge: > > >>> > > >>> Break circular dependencies between the 2 projects > > >>> Remove redundant copy/pasted code > > >>> Simplify build and CI > > >>> Reduce friction on changes that span both projects > > >>> Simplify the CDC implementation > > >>> > > >>> > > >>> Outstanding questions and observations that came up: > > >>> > > >>> Do we merge one repository into the other? Or do we create a new > project and bring them both in? > > >>> What do we do about JIRA? Leave separate or combine? > > >>> What do we do with open issues and PR's in github? > > >>> We'll need to thoughtfully update CI (github + circle) since we're > right at the limit on the free tier on both projects > > >>> What do we do about existing deprecated repositories > (cassandra-analytics and/or cassandra-sidecar)? > > >>> We'll need to update our release process > > >>> > > >>> > > >>> Other observations or questions welcome, as are thoughts on the > entire process, on the outstanding questions, etc. > > >>> > > >>> Looking forward to the discussion everyone. > > >>> > > >>> ~Josh > > > > >
