Description of problem: after updating to latest a Mga5 i586 (6 DE in 1 install),using a Mga6-sta1 USB as source, partway through an upgrade of Mga5 i586 (6 DE in 1 install)a popup advising "Error: 'script' failed for confg...", and "Error: 'script' failed for gcc...." is displayed. closing the popup presents a window advising that 800+ packages failed to install Version-Release number of selected component (if applicable): Mageia-6-sta1-i586-DVD.iso DATE.txt: Mon Jun 20 23:38:16 CEST 2016 How reproducible: Steps to Reproduce: 1.create an Mga5 i586 installation with 6 main DE. update to latest 2.start upgrade from above .iso via USB 3.wait for error notice to pop up
Keywords: (none) => 6sta1
re-running the install again using the failed installation partitions, but using non-free net install, presented again a popup that advised "Error:'script' failed for epiphany. closing the popup, the installation proceeded to the config window with no further issues set bootloader, rebooted successfully. of note KDE4 presented in the sddm DE options
Summary: [6sta1]partway through upgrade from Mga5 to Mga6 "Error: 'script' failed for confg...." => [6sta1]partway through upgrade from Mga5 to Mga6 "Error: 'script' failed for GConf2-3.2.6-8mga5.i586...."
correction. "Error: 'script' failed for > confg...", and "Error: 'script' failed for gcc...." is displayed. should be "Error: 'script' failed for GConf2-3.2.6-8mga5.i586" "Error: 'script' failed for Glibc-6.2.22-19.mga5.i586"
CC: (none) => thierry.vignaud, tmbAssignee: bugsquad => pkg-bugs
That's not an installer issue but packaging issues
CC: thierry.vignaud => (none)Component: Installer => RPM Packages
on retry overnight i586 6in1 from image checked for updates. none used net installer non-free boot.iso from usb. set http://distrib-coffee.ipsl.jussieu.fr/pub/linux/Mageia/distrib/cauldron/i586 as source again, same popup, Glibc + Gconf2, also this time with "mate-terminal. as the rest is not displayed, am unable to give versioning #
upgrading M5-x86-64 6in1 DE to mga6 (estimating @ +24hrs) same errors "Error: 'script' failed for GConf2-3.2.6-8.mga5.x86_64" "Error: 'script' failed for Glibc-6:2.22-19.mga6.x86_64" "Error: 'script' failed for epiphany-3.20.1-1.mga6.x86_64" "Error: 'script' failed for mate-terminal-1.14.0-1.mga6.x86_64" corrections of version numbers also
Only script for GConf2 from mga5 that could be firing, but *shouldn't* be is: %triggerpostun -- GConf2 < 2.22.0-4mdv GCONF_CONFIG_SOURCE=`%{_bindir}/gconftool-2 --get-default-source` %{_bindir}/gco nftool-2 --makefile-install-rule %{_sysconfdir}/gconf/schemas/*.schemas > /dev/n ull are we sure rpm isn't doing something wrong here?
CC: (none) => thierry.vignaud
checked again last night. upgrade from M5 (6in1 DE) to M6 using boot.iso and local HDD i586.iso as install medium. failed with the above 4 (+1 more below mate-terminal, but unable to read it as window is too small) :(
same for x86-64 upgrade
Priority: Normal => release_blockerBlocks: (none) => 15527
Keywords: (none) => 6RC
Still valid for 6RC rnd3 i586
epiphany and mate-terminal are both calling update-alternatives --remove in %post.
I hope http://gitweb.mageia.org/software/rpm/perl-URPM/commit/?id=e134398e5535ddf0e4fbda067eeffd7581a9b339 means we'll get some more details about the scripts that failed next time this happens. @ Thierry: Are there retrievable logs after such a failed upgrade that could be used to debug here?
Status comment: (none) => Scripts fail and block upgrade, should be debugged and fixed asap
Nope, that's when checking the transaction (for things that would make it to abort). All logs are in /root/drakx/report.bug.xz
Would be great to have this bug tested again either via a net install or when the new 6 sta2 ISOs land in QA testing. Thanks :)
(In reply to Rémi Verschelde from comment #13) > Would be great to have this bug tested again either via a net install or > when the new 6 sta2 ISOs land in QA testing. Thanks :) And report.bug.xz attached to the bug report.
Status comment: Scripts fail and block upgrade, should be debugged and fixed asap => Scripts fail and block upgrade. Needs someone to debug them.
From Thomas on IRC: <tmb> well, wothout looking closer right now the fix for 18890 might need to be fixed in mga5 (as mga6 package might already be fixed)... <tmb> or just add an upgrade trigger in mga6 doing the stuff the mga5 package uninstall fails to do...
(In reply to David Walser from comment #10) > epiphany and mate-terminal are both calling update-alternatives --remove in > %post. And GConf2 is calling update-alternatives --remove-all in the %triggerpostun which shouldn't be getting called, but I guess maybe it is, and --remove-all is not a valid option, --remove is. I guess that call should be changed to --remove gconftool /usr/bin/gconftool-2
I installed Mga6 this past weekend using the last Mageia-6-RC-i586-DVD.iso from August and there were No "failed script" errors. If/when new Mga 6 iso are created I will test again.
CC: (none) => cae
Created attachment 8740 [details] install report.bug x86_64
Summary: [6sta1]partway through upgrade from Mga5 to Mga6 "Error: 'script' failed for GConf2-3.2.6-8mga5.i586...." => [6sta2]partway through upgrade from Mga5 to Mga6 "Error: 'script' failed for GConf2-3.2.6-8mga5.i586...."Keywords: (none) => 6sta2
upgrade failed for x86_64. report.bug captured \o/
(In reply to ben mcmonagle from comment #18) > Created attachment 8740 [details] > install report.bug x86_64 I see file conflicts in KDE packages and LXDE packages, but not script failed errors. Please file a new bug for this and assign to the KDE team and CC ns80.
CC: (none) => luigiwalser
Created attachment 8742 [details] photo of :ERROR`script' (In reply to David Walser from comment #20)
%triggerin(postfix-1:2.10.3-5.mga5.x86_64) scriptlet failed, exit status 127 %triggerin(systemd-217-11.2.mga5.x86_64) scriptlet failed, exit status 127 %triggerpostun(GConf2-3.2.6-14.mga6.x86_64) scriptlet failed, exit status 2 %post(mate-terminal-1.14.0-1.mga6.x86_64) scriptlet failed, exit status 2 I think I just fixed the GConf2 and mate-terminal ones. The exist status 127 means command not found, and those two are the ones in the picture showing as "glibc" because they are triggerin's on glibc. It seems nonsensical that it's erroring, as those triggers come from the packages that provide the command that they're calling. %triggerin -- glibc setup nss_ldap nss_db nss_wins nss_mdns # Generate chroot jails on the fly when needed things are installed/upgraded %{_sbindir}/postfix-chroot.sh -q update %triggerin -- glibc # reexec daemon on self or glibc update to avoid busy / on shutdown # trigger is executed on both self and target install so no need to have # extra own post if [ $1 -ge 2 -o $2 -ge 2 ] ; then %{_bindir}/systemctl daemon-reexec 2>&1 || : fi
Anyway, still please file a bug for the file conflicts, as I asked before.
(In reply to David Walser from comment #23) > Anyway, still please file a bug for the file conflicts, as I asked before. Hi Ben. Has the bug report for file conflicts been file as asked? Is this bug report still valid or are the "script failed" issues gone?
He did file the bug, it was Bug 19914. Nicolas already fixed the KDE conflict issues. I'm not sure if Nicolas fixed the LXDE ones.
(In reply to David Walser from comment #25) > He did file the bug, it was Bug 19914. Nicolas already fixed the KDE > conflict issues. I'm not sure if Nicolas fixed the LXDE ones. Hi, The LXDE conflict issue comes from the split of lxsession package into lxpolkit and lxsession packages. If I add "Conflicts: lxsession < 0.5.3" in "%package -n lxpolkit", do you think the problem will be solved? Best regards, Nico.
CC: (none) => nicolas.salguero
(In reply to Nicolas Salguero from comment #26) > (In reply to David Walser from comment #25) > > He did file the bug, it was Bug 19914. Nicolas already fixed the KDE > > conflict issues. I'm not sure if Nicolas fixed the LXDE ones. > > Hi, > > The LXDE conflict issue comes from the split of lxsession package into > lxpolkit and lxsession packages. If I add "Conflicts: lxsession < 0.5.3" in > "%package -n lxpolkit", do you think the problem will be solved? > > Best regards, > > Nico. That sounds like the correct solution.
(In reply to David Walser from comment #27) > That sounds like the correct solution. Done then.
So can we close this bug report now?
I have no objection. But I am waiting for a new Classical Install.iso to test with. If issue is still apparent, I will re-open this report
Closing then, thanks!
Status: NEW => RESOLVEDResolution: (none) => FIXED
Created attachment 8850 [details] new report.bug.xz upgrading a 5.1 i586 (fully updated) using a local HDD .iso of Mageia-6-sta2-i586-DVD.iso DATE.txt: wed Jan 11 09:54:21 CET 2017 using the boot.iso installer resulted in an ÃRROR:`script' failed for glibc..... attaching the new report.bug.xz
:(
Status: RESOLVED => REOPENEDResolution: FIXED => (none)
(In reply to ben mcmonagle from comment #32) > Created attachment 8850 [details] > new report.bug.xz > > upgrading a 5.1 i586 (fully updated) using a local HDD .iso of > > Mageia-6-sta2-i586-DVD.iso > DATE.txt: wed Jan 11 09:54:21 CET 2017 > > using the boot.iso installer > > > resulted in an ÃRROR:`script' failed for glibc..... > > attaching the new report.bug.xz I filed bug 20111 for that issue, because this report becomes too long and because the original issue of this report got fixed.
Status: REOPENED => RESOLVEDCC: (none) => marja11Resolution: (none) => FIXED
Blocks: 15527 => (none)