Don't sure if this is right place to post, but since I'm experimenting with catalyst
maybe someone will share some thoughts and suggests additional reading.

What I'm trying to do:

1. Create small, task-oriented stage4 for fixed hardware configuration
2. Deploy it to several machines (now looking for deployment variants and installers)
3. Maintain it (updates)

The most interesting and challenging part is (3). There are my thoughts and 
associated questions:

I want special build host, where all binary targets are built. Its setup
should be easily repeatable by other developers (i.e. every task should be scripted,
without manual tweaks like 'chroot here and patch/emerge this'). Catalyst with
specs and overlays seems fine to me.

- Which parts of host system affect on resulting target?

- I probably do not want update portage tree and world on build host very often,
  but want newer snapshots and seed stages to build targets. Is it possible?

I want incremental selective binary updates to be pushed on nodes. I.e. generate
stage4 once, and push only changes for a long period of time. Then repeat.

- Is it possible with catalyst to produce icremental binary updates (i.e. feed it with
  new snapshot and take set of updated packages as a result)? To me seems it is not.

- Is portage tree required on nodes in order to install binary packages (never used them)?

- My first plan is to unpack stage4, put fresh tree snapshot, clean PKGDIR and update world.
  Then push everything from PKGDIR to nodes (still don't know how to manage configs).

- What about sequential stage4 generation without purging cache?

- Get rid of portage on nodes and push binary diffs (or rsync).

And small questions about catalyst internals. Is it possible to use some functions manually?
For example, I want to mount-bind required parts (/proc, /dev, snapshot, distfiles...),
chroot to storedir/tmp/default/stageX, do or test something manually and then exit and unmount.
Is there are any handy scripts to do this?  Manual catalyst execution step-by-step?
Also it will be good to just look on emerge -pv output, or list URLs which catalyst
tries to fetch (for manual fetching and offline operations).


Any thoughts?  How do you guys organize things like this?
-- 
gentoo-catalyst@g.o mailing list