Re: something seriously wrong with my bookworm install
On Wed 26 Jul 2023 at 15:03:55 (-0400), gene heskett wrote:
> On 7/26/23 10:32, David Wright wrote:
> > On Wed 26 Jul 2023 at 10:07:34 (-0400), gene heskett wrote:
> > > And since bookworm has shut down, or moved, all the logs that might
> > > keep track of this, I'm lost. What can I do to trace or fix the reason
> > > for this denial of service? I know zip about your new ACL stuff if
> > > thats even involved. IDK. Is there a special version of chown for
> > > raid arrays?
> > > And where the heck are the logs, they should be in the /var directory
> > > of the drive its booted from which s/b /dev/sda, but df thinks its:
> > > now /dev/sda, last boot it was /dev/sdb, so much for UUID's. I have
> > > not moved any cables. One is a 500G samsung 860 SSD, the other a 1t
> > > Samsung 870 SSD.
> > > An ls of /var/log:
> > > gene@coyote:~$ ls /var/log
> > > alternatives.log apache2 boot.log boot.log.2 boot.log.4 btmp
> > > cups dpkg.log.1 faillog gdm3 journal private
> > > runit sddm.log wtmp
> > > alternatives.log.1 apt boot.log.1 boot.log.3 boot.log.5
> > > btmp.1 dpkg.log exim4 fontconfig.log installer lastlog
> > > README samba
> > > speech-dispatcher
> > >
> > > So where are syslog and dmesg?
> >
> > I take it you didn't bother to read §5.1.7 of the bookworm Release Notes.
>
> Not yet David. I was forced to install bookworm after an update wiped
> out the validity of my pw. So I went to another machine on my net and
> downloaded and wrote a dvd with the bookworm netinstall. so at no
> time was I presented with an opportunity to read the release notes.
It's a long time since the dog ate my homework. Anyway, dmesg
hasn't gone anywhere as it's in util-linux, a Required package.
Typing journalctl will give you the journal in full AIUI, but
you can add -b to limit it to this boot. man journalctl has more
options; the Release Notes prefer -e for showing the last 1000 lines.
> This to me, is not a toy, And that's not the first time my pw has
> been invalidated by an update. No root pw has ever been set since
> wheezy, so I had no choice but to install either bookworm or wheezy as
> I seem to have misslaid the diskj for the in betweens.
I don't know why you don't use a root password. It's never seemed
a sensible choice to me, so I always say yes during installation.
> Wheezy thru buster has been good to me, bullseye was troublesome but
> fixable, now bookworm is a disaster. The applications I use daily act
> like they don't have write perms to storage I own lock stock and
> barrel. The logs I might use to troubleshoot this myself are gone,
> and you want to know if I read the release notes? So I come here
> asking for help with all the changes and catch it for not reading the
> notes i never had the chance to read.
You're an experienced user. I think you've been running bookworm
(or trying to) for about three weeks? It takes four clicks from
the Debian home page to read them (or download them as a PDF).
And however wheezy was, once it settled down, installing it led
to long threads here through 2015, about how broken it was.
> I ask how to do an fsck on it which might fix whatever changes
> bookworm has done to ext4. One possibility is to comment that line
> that mounts it out of fstab and reboot. But alluding to that before
> has been ignored.
It's still as it was in:
https://lists.debian.org/debian-user/2017/10/msg00399.html
ie forcefsck on the kernel line when you boot, though I'm lazy,
and understand that forcefsck's Sunday name is fsck.mode=force.
Cheers,
David.
Reply to: