Red Hat Bugzilla – Bug 1208296
coredump when running do_transaction in a separate process with https repos
Last modified: 2015-12-02 12:49:20 EST
vpodzime discovered a problem with lorax while using it with a copr repo. I've simplfied a reproducer and tracked it down to using https repos. I'm not sure if this is an actual bug, or just a limitation of how SSL is setup by dnf/hawkey/rpm/whatever. lorax uses the python2 multiprocessing module to run the do_transaction() in a separate process. If one or more of the repos used is https this will coredump down in librpm. I've attached the traceback logged by systemd-coredump as well as the coredump file.
Created attachment 1009866 [details] script to demonstrate the problem You'll have to edit this to change the proxy settings or remove them.
Created attachment 1009867 [details] backtrace from systemd-coredump
Running on a F21 system with: dnf --version 0.6.4 Installed: dnf-0:0.6.4-1.fc21.noarch at 2015-02-23 15:25 Built : Fedora Project at 2015-02-09 12:56 Installed: rpm-0:4.12.0.1-5.fc21.x86_64 at 2015-03-23 02:38 Built : Fedora Project at 2015-03-03 17:32
And you can find the coredump here - https://bcl.fedorapeople.org/coredumps/core.dnf-multiproces.0.2333da07ab304a6baa9791c472c70c1c.2277.1427923406000000.xz
It crashes on my system too. Whats the reason to run do_transaction in multiple processes? There's still dnf lock above rpmdb. I thought that you run downloading of packages in one process and installation in another.
That's the way Ales wrote it -- the lorax code was modeled on the dnfpayload module from anaconda. In reality we just need to be able to feed the callback information to the UI while do_transaction is running, so a non-multiprocessing thread works fine for that (vpodzime wrote a patch to do that).
(In reply to bcl@redhat.com from comment #6) > That's the way Ales wrote it -- the lorax code was modeled on the dnfpayload > module from anaconda. > > In reality we just need to be able to feed the callback information to the > UI while do_transaction is running, so a non-multiprocessing thread works > fine for that (vpodzime wrote a patch to do that). That however has the problem with rpm doing chroot() when processing RPMs. Because of that we need to run the transaction in a separate process, thread is not enough, unfortunately. However, that doesn't apply to the downloading phase for which threads are okay.
From today rpm discussion on IRC: - rpm don't want to do fakechroot instead to ensure scriplets are executed in right path - there could be function in rpmlib which will spawn another process, chroot it and IPC to the parent (there could be issues with other libs having problem with fork)
In lorax we don't really need threads or processes, it isn't doing anything else. This patch drops all that extra stuff: https://github.com/rhinstaller/lorax/pull/50 Should we just move this back to lorax? This fixes for me.
This message is a reminder that Fedora 21 is nearing its end of life. Approximately 4 (four) weeks from now Fedora will stop maintaining and issuing updates for Fedora 21. It is Fedora's policy to close all bug reports from releases that are no longer maintained. At that time this bug will be closed as EOL if it remains open with a Fedora 'version' of '21'. Package Maintainer: If you wish for this bug to remain open because you plan to fix it in a currently maintained version, simply change the 'version' to a later Fedora version. Thank you for reporting this issue and we are sorry that we were not able to fix it before Fedora 21 is end of life. If you would still like to see this bug fixed and are able to reproduce it against a later version of Fedora, you are encouraged change the 'version' to a later Fedora version prior this bug is closed as described in the policy above. Although we aim to fix as many bugs as possible during every release's lifetime, sometimes those efforts are overtaken by events. Often a more recent Fedora release includes newer upstream software that fixes bugs or makes them obsolete.
Fedora 21 changed to end-of-life (EOL) status on 2015-12-01. Fedora 21 is no longer maintained, which means that it will not receive any further security or bug fix updates. As a result we are closing this bug. If you can reproduce this bug against a currently maintained version of Fedora please feel free to reopen this bug against that version. If you are unable to reopen this bug, please file a new report against the current release. If you experience problems, please add a comment to this bug. Thank you for reporting this bug and we are sorry it could not be fixed.