Please ensure Jenkins isoworkers still can run the Tails test suite once they are upgraded to Trixie

Trixie defaults to a tmpfs on /tmp whose size is 50% RAM. This may impact any system that runs the Tails test suite, which uses lots of space in /tmp/TailsToaster by default. This issue is specifically about Jenkins isoworkers. Please ensure they're ready for this change, so that test suite runs don't all start failing as soon as you upgrade these VMs to Trixie.

At least on isoworker4 I see /tmp/TailsToaster is on the root filesystem, so if it got upgraded to Trixie as-is, things would start breaking.

I'll list a few options I can think of, I don't have a horse in this race, please do whatever is best for TPA as long as things keep working for Tails developers :)

disable the "tmpfs on /tmp" change, before or when you upgrade to Trixie (can be done proactively with a symlink to /dev/null IIRC, which will be a no-op until the upgrade)
give isoworkers a new virtual disk and mount it on /tmp/TailsToaster (we have code to manage it in tails::profile::jenkins::isotester and tails::profile::jenkins::tester); quite wasteful in terms of disk space but might allow fancy I/O optimizations
configure the test suite to use another temporary directory, backed by the root filesystem; that's done with the TMPDIR setting in the YAML test suite config deployed on all isoworkers; if you go this way, please ensure that directory is cleaned up on boot (examples: https://gitlab.tails.boum.org/tails/tails/-/merge_requests/1866, https://gitlab.tails.boum.org/tails/puppet-tails/-/merge_requests/161/diffs + follow-up fixes)