CentOS - Dec 2018 - Weird problems with CentOS 7.6 1810 installer

Hi,

I just updated my installation media for CentOS. I have a few sandbox
PCs in my office, and I'm testing CentOS 7.6 1810 on them. There seem to
be a few issues with the CentOS 7.6 1810 DVD.

Checked DVD integrity on startup : OK.

First attempt : installer froze on root password dialog.

Second attempt : installer froze on dependency calculation.

Third attempt : installer froze on network interface definition.

Similar behavior on two different machines.

On the same machine, the minimal DVD seems to work OK.

I could manage to switch to another virtual console, and I have the
following kernel log messages:

WARNING: kernel:perf: interrupt took too long

Any idea what's going on here?

I should add that all the previous CentOS 7 installation DVDs worked
perfectly on these PCs (Dell Optiplex 330).

Cheers,

Niki
-- 
Microlinux - Solutions informatiques durables
7, place de l'?glise - 30730 Montpezat
Site : https://www.microlinux.fr
Blog : https://blog.microlinux.fr
Mail : info at microlinux.fr
T?l. : 04 66 63 10 32

On Wed, 5 Dec 2018 at 02:02, Nicolas Kovacs <info at microlinux.fr>
wrote:
>
> Hi,
>
> I just updated my installation media for CentOS. I have a few sandbox
> PCs in my office, and I'm testing CentOS 7.6 1810 on them. There seem
to
> be a few issues with the CentOS 7.6 1810 DVD.
>
> Checked DVD integrity on startup : OK.
>
> First attempt : installer froze on root password dialog.
>
> Second attempt : installer froze on dependency calculation.
>
> Third attempt : installer froze on network interface definition.
>
> Similar behavior on two different machines.
>
> On the same machine, the minimal DVD seems to work OK.
>
> I could manage to switch to another virtual console, and I have the
> following kernel log messages:
>
> WARNING: kernel:perf: interrupt took too long
>
> Any idea what's going on here?
>
> I should add that all the previous CentOS 7 installation DVDs worked
> perfectly on these PCs (Dell Optiplex 330).
>
Well the Captain Obvious answer is something hardware related is
taking too long while the system is working out sample loads.
https://bbs.archlinux.org/viewtopic.php?id=187636 It is supposedly an
informational versus the reason the system died.

How much memory do the systems have, what is its CPU, and how many
packages are you installing? [The Dell Optiplex 330 is from 2007 and
came with several different layouts so it might need info.] Finally
does the install work if you only choose what the minimal would
install?


> Cheers,
>
> Niki
> --
> Microlinux - Solutions informatiques durables
> 7, place de l'?glise - 30730 Montpezat
> Site : https://www.microlinux.fr
> Blog : https://blog.microlinux.fr
> Mail : info at microlinux.fr
> T?l. : 04 66 63 10 32
> _______________________________________________
> CentOS mailing list
> CentOS at centos.org
> https://lists.centos.org/mailman/listinfo/centos


-- 
Stephen J Smoogen.

On 12/4/18 11:03 PM, Nicolas Kovacs wrote:
> I could manage to switch to another virtual console, and I have the
> following kernel log messages:
> WARNING: kernel:perf: interrupt took too long

Can you run "dmesg" and find out if there's any more info?? Often
there
will be a kernel stack trace that provides hints as to which module was 
handling the interrupt, and therefore which device or driver is at fault.

>
>
>
> Commands line options:
rd.debug rd.udev.debug systems.log_level=debug

That willl be incredibly verbose, and slows things down a lot, so in the
off chance there's a race, you might get different results. But if not, the
log should contain something useful.

I like the hypothesis about mdadm metadata version 0.9, however that's
still really common on RHEL and CentOS. It was used for /boot prior to
about Fedora around Fedora 24 maybe?

It could also be a dracut bug, since that's what's largely responsible
for
assembly. But then it can be confused by a change in udev rules. :-D

However, going back to mdadm 0.9 metadata, that version should only be
kernel auto detected. Theoretically, it gets activated before dracut gets
involved. 1.x versions have no kernel autodetect, instead it happens in
dracut (by calling mdadm to assemble and run it).

Oh, is this really dm-cache, not lvmcache? That might be a source of
confusion, if there isn't lvm metadata present to hint at LVM for proper
assembly. Of course, lvmcache still uses device mapper, but with LVM
metadata.

Anyway is quite an interesting, and concerning problem.


Chris Murphy