Red Hat Bugzilla – Bug 240138
ctrl-c/SIGINT yum vs. rpm, checkSignals() in rpm kills yum
Last modified: 2009-01-20 16:44:01 EST
Description of problem:
when yum catches a single ctrl (for example to switch to a faster mirror for
downloading) it finishes the download, but does not install the package. Yum
should die completely on two ctrl-c's, but if the application continues, then it
should finish all steps.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. install package from slow mirror
2. hit ctrl-c once to change mirrors
3. watch yum fetch package, but not install it.
yum fetches package but does not install.
yum should switch to new mirror, and continue initial action of installing package.
This behavior also occurs in every version of yum in fedora up to the -testing.
I've not tested with the yum in the fc development tree.
I was using the development tree till yesterday, the behavior is the same.
Maybe Ctrl-\ (back slash) should kill yum instantly if ctrl-C will be used to
switch mirrors (quite useful feature, I use it often and when it finishes
downloading I run the command again - the packages are already downloaded).
Well, the code states that 2 quick ctrl-c's should kill yum, and it does. I
would keep ctrl-c as a kill, as this is standard, but either allow yum to finish
installing if it doesn't die, or use another key combination to switch mirrors.
This should be better as of yum 3.2.5 if you have rpm 18.104.22.168 or later
I don't see 3.2.5 in RHEL5 or in the RHEL5 beta channel on rhn... where should I
be looking, other than at linux.duke.edu
Yeh, that's why I re-opened it. Hopefully we'll get 3.2.5 or so in for RHEL-5.2.
I'm guessing Jeremy thought it was a Fedora bug.
sounds good. if there's a beta package for RHEL5 later on, I'll happily test.
This request was evaluated by Red Hat Product Management for inclusion in a Red
Hat Enterprise Linux maintenance release. Product Management has requested
further review of this request by Red Hat Engineering, for potential
inclusion in a Red Hat Enterprise Linux Update release for currently deployed
products. This request is not yet committed for inclusion in an Update
there is new yum (3.2.8-9.el5) in the 5.2 Beta channels on the RHN. It would be
really great if you could test it because it is better to have more eyes on it.
Ok, yeh I can confirm the bug here ... the problem is that it's an rpm thing.
From the yum side we go into:
...and we never come out. There's no customer stuff hanging on this (and RHN
doesn't do mirrors), and I'm assuming panu doesn't want to fix this at this
point. So we should probably just remove the bug from the yum errata.
I'll try and find out if we know when this got fixed in Fedora.
Moving to 5.3 (and we may need to reassign to rpm)
Well it's still on the yum component, but I have a pretty big suspicion that
this is the same underlying bug as rhbz#429737 ... and so is a yum bug.
Given that we don't have multiple servers, we only really need to make sure
that this works in what will be the 5.3 errata version ... and probably the
easiest thing to do is make sure the fix
(a726a9d7652fff46000b5ac19de50416ef342719) is tested against the 5.2 yum and/or
the latest upstream yum doesn't have the problem.
Yeh, that's the one bug#442232 ... not sure why I like the other one so much :).
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.