Commit 275b2444 authored by Antonio Terceiro's avatar Antonio Terceiro

cleanup TODO list: remove items that are DONE

parent 4f1b5e70
[X] use Packages/Sources indexes from chroot instead of the ones in the system
[X] automate creation/maintainance of chroot
[X] wrapper script to be called from cron
[X] updates chroot
[X] automate deployment (package it)
[X] unify status and log directories
[X] arrange definitive^Winitial hosting
: ssh
[_] setup https
[_] first adt-run execution will sometimes hang
: waiting for gpg key generation due to low entropy
[_] add timeout to guard against unbehaving test suites
[_] support for blacklisting packages
[_] blacklist pyzmq: for some reason it is crashing the entire run
[X] add latest.log symlink as well
[_] add a postrm to remove all traces on package purge
[_] remove data directory
[_] remove schroot setup
[X] add option to force reprocessing all packages
: (say when the system itself changes)
[_] figure out how to deal with test suites that leave garbage behind
: schroot sessions are left behind. some of them I can just terminate if
: `schroot --end-session --chroot $SESSIONID`; others won't go away because
......@@ -25,14 +14,8 @@
: leftover /usr/bin/gdomap daemon running (didn't figure out which package
: leaves that yet).
: maybe the cleanest solution is going full virtualization with either kvm
: or lxc instead of schroot.
[X] add status chart showing run duration over time
[X] take base system into consideration (?)
[X] re-process all packages when the base system changes
: maybe it's better to re-run all packages periodically instead
[X] store the state of the base system
[X] log differences in the base system
: maybe the cleanest solution is going full virtualization with kvm or at
: least a lxc container instead of chroots despite how good schroot is.
[_] publish atom feeds
[_] global status changes: daily, weekly, monthly
: which packages started to tail their tests, which packages started
......@@ -45,38 +28,11 @@
[_] generate a dd-list of failing packages
[_] graceful degradation for user with JS disabled
: or generate only static HTML
[X] record reason for the test being run
: <zack> jcristau: does satisfy
: release team need about knowing which packages pass/fail
: autopkgtest?
: <jcristau> i think we need to be able to know when something changed from
: pass to fail
: <nthykier> zack: terceiro: I think we need to know why the test was run as
: well (ie. "new version of the package" vs. "retest because it
: is an rdep of X, which got a new version")
: [...]
: <nthykier> terceiro: zack: Re: the output - I
: already mentioned that I believed we needed the "blame" in the
: output (i.e. which package is to blame for this failure).
: Come to think of it, it will probably need to "remember" the
: blame across fails so the blame is not reassigned)
[_] procenv
: <doko> and could you run procenv after setting up the chroot?
: <doko> my experience is that a lot of time goes into setting up the
: environement and understanding the env
[X] display autopkgtest version in the logs
[X] add tmpfail result
: differentiate a problem in the environment (mirror failure, tested skipped
: because of missing features in testbed etc) from actual failures.
: add better handling of adt-run exit statuses
[_] setup a BTS !!
[X] move command line processing to lib/
[X] packaging: make debci $HOME a directory that _will_ be available inside the chroot
: for context see
: - <>
: - <>
[_] also test packages from experimenal
[X] upgrade autopkgtest to the version in jessie (2.7)
[_] list largest log files
[_] list longest duration between test jobs
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment