Hi,

I have some problems understanding the breaks-testbed restriction:

  breaks-testbed
    The test, when run, is liable to break the testbed system. This
    includes causing data loss, causing services that the machine is
    running to malfunction, or permanently disabling services; it does
    not include causing services on the machine to temporarily fail.

AIUI the testbed system is a throwaway chroot/container/virtual machine/...
So how can breakage (e.g. rm /bin/ls) be persistent? Or is it possible
to run subsequent tests in the same testbed instance?
Or does that mean the test can affect the host system? (E.g. from a
chroot with root access killing all processes with even PIDs, including
processes on the host.)
BTW, what does "machine" denote in the quoted paragraph?

Or am I on the wrong track here?


Andreas

as an optimization autopkgtest reuses testbeds across tests if the
dependencies of the later test are a superset of the dependencies of the
earlier ones, unless breaks-testbed is specified, in which case the
testbed is discarded after the test.

that depends on specifics of the tests and of the virtualisation backend
you are using.

note that the very next paragraph says:

    When this restriction is present the test will usually be skipped
    unless the testbed's virtualisation arrangements are sufficiently
    powerful, or alternatively if the user explicitly requests.

the host machine.

Do you have a wording suggestion to make it more clear?

Hi,


So, we discussed that we could document reuse of the testbed a bit
better. But then, during that discussion, LP 2058040 [1] was referenced
and we discussed we might want to reconsider how we're dealing with reuse.

Paul

[1] https://bugs.launchpad.net/ubuntu/+source/autopkgtest/+bug/2058040