Re: Reducing apt's memory footprint (on small boxes)
On Sun, Feb 14, 2021 at 1:42 PM Christoph Biedl wrote:
> The packages indexes became that big they no longer fit into memory. In
> September 2015, there was a suggestion to create subsets of a release
I seem to remember that Emdebian used that solution too.
> but I objected it will be more or less impossible to create them without
> breaking some builds or installations for unresolved dependencies.
This should be easy to avoid, by using apt on a larger system to get
the list of packages to install, and then creating a subset of
Packages based on the resulting list. I believe the germinate package
(that Ubuntu uses for their images) does something like this, but
using the apt Python bindings instead of `apt install
--download-only`.
> Now, from my experience: Every good idea that I believe to have found
> turned out had been prosoped earlier by somebody else and/or was not as
> good as it initially seemed. So, where's the actual catch in that setup?
I believe that the Emdebian project shut down because hardware with
very little RAM/flash became much rarer.
> Or is that something that might be of broader interest and possibly even
> become part of a Debian release?
I think that this could be useful to a subset of Debian users,
possibly including embedded hardware and low-RAM cloud/VPS users.
> Of course there is the problem of defining which packages should be
> included in such a "debian-narrowed"¹.
I suggest that instead the desired packages should be provided by
users and the service should dynamically calculate the package install
sets based on the seed packages. The service could calculate the N
most popular package install sets after every mirror push and
dynamically recalculate the ones with outdated or missing cache
information when users request them.
--
bye,
pabs
https://wiki.debian.org/PaulWise
Reply to: