http://bugzilla.suse.com/show_bug.cgi?id=953987
http://bugzilla.suse.com/show_bug.cgi?id=953987#c5
Leslie Satenstein changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |lsatenstein@yahoo.com
Flags|needinfo?(leslie.satenstein |
|@gmail.com) |
--- Comment #5 from Leslie Satenstein ---
I reinstalled. During the installation I watched /mnt/var/log/messages
Th3 following was run at least 95 times with long long delays. The log message
was...
mounted-tests running subtests /usr/lib/os-probes/mounted/90linux-distro
I presumed wrongly that the installation logs would be preserved within the
target /var/log/installation or some /var/log directory. How can you do a
proper trouble shooting without the installation logs?
Here is some info to aid in the debugging.
Normal time to Kernel Installation step (233 files to go on display screen)
Kernel install took 20 minutes
boot at end of installation took 1 hour or more with cpu stuck at 96%
(I left to do shopping)
Let me tell you wnat my system has: All drives formatted gpt. This system has a
regular bios (non uefi). Dual core Intel e7300cpus 8gigs ram 900mhz.
device contains
/dev/sda windows 10 for 500gigs and ntfs data for 500gigs
/dev/sdb ChapeauLinux (Fedora23 Remix test) and spare global partitions
/dev/sdc KororaLinux (Fedora23 Remix test) and spare partitions
/dev/sdd openSUSE-Leap-42.1-DVD-x86_64.iso on 155gigs, with iso sha256sum of
8576e84822cdbe566bf551e28a169fc028229831eba9f07a4c1f84302c5ddb09
+ spare partitions
/dev/sde SSD with ChapeauLinux(Fedora 22/Fedora23) daily. This is an SSD drive
There is a loop with y2base(100%) followed by grub2-mount(100%)
Within the messages file (as this was not preserved after the installation),
grub 90linux message* appeared 95 times. There is a bug, a real bug, Two
hours of cpu time is enough
Why does y2base and grub2-mount have to be called 95+ times for an
installation.
Can't the system remember the values obtained from the first execution?
Please, at end of installation, before the boot, can't you copy the
installation log directory to the physical /var/log ? After all, the final
/var directory was created at the beginning of the boot process and can be
written to. (Fedora's anaconda copies its logs to /var/log/annaconda. Suse
should do likewise).
Alternately, do not timeout and perform an automatic reboot and end of run. Let
the user trigger the reboot. (or ask if an automatic reboot should be omitted)
==================================================
Bash script
date -u >time.log
time os-prober >> time.log 2>>&1
date -u >> time.log
Thu Nov 12 19:41:37 UTC 2015
/dev/sda1:Windows 8 (loader):Windows:chain
/dev/sdb6:Chapeau release 23 (Armstrong):Fedora:linux
/dev/sdc6:Korora release 23
(Coral):Fedora1:linux:btrfs:UUID=5f629c52-1717-4d68-bb3b-61b36f8129bb:subvol=root
/#dev/sde4:Chapeau release 23 (Armstrong):Fedora2:linux
real 15m26.457s
user 0m0.564s
sys 0m0.412s
Thu Nov 12 19:57:04 UTC 2015
====================================================
--
You are receiving this mail because:
You are on the CC list for the bug.