[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: trying to get bookworm net going on an rpi4



On Sun, Jan 23, 2022 at 05:26:17PM -0500, gene heskett wrote:
> On Sunday, January 23, 2022 4:28:56 PM EST Andrew M.A. Cater wrote:
> > On Sun, Jan 23, 2022 at 04:08:34PM -0500, gene heskett wrote:
> > > Greetings folks;
> > > 
> > > Let me say first that I'm around 150 miles from any ipv6 capable
> > > network, just to get that out of the way.
> > > 
> > > Its booted to a text login and the first thing I did was import a
> > > saved copy of the hosts file, and added rpi4-20220122.coyote.den
> > > with an alias of rpi42 to that file. Then I edited
> > > /etc/network/interfaces.d/eth0 to set a static address, and added a
> > > dns-nameserver 192.168.xx.1 line to it, my router and gateway, and a
> > > second additional line to set /proc/sys/ net/.../eth0/disable_ipv6
> > > to a 1. And I've checked, disable_ipv6 is now a 1.
> > > 
> > > But that's not enough, ip a, after a /e/i/networking restart, still
> > > shows its active, so I assume its sending its dns request out the
> > > ipv6 portal and my router is a dummy load for anything ipv6. So I
> > > can ping all the local machines, but cannot look up yahoo.com.
> > > 
> > > So question one:
> > > So what is the official, works in bookworm every time, way to totally
> > > kill ipv6, making it use ipv4 for everything?
> > > 
> > > This ought to be in a FAQ somewhere. Except everybody has their own
> > > method. :(
> > > 
> > > And question 2:
> > > I noticed also, there is not a first user, can I assume the first
> > > time
> > > adduser gets run it will assign him/her/it as user 1000 AND will put
> > > that user into sudoers?
> > > 
> > > Again, this is the 20220121 bookworm.
> > > 
> > > Thanks all.
> > > 
> > > Cheers, Gene Heskett.
> > 
> > Gene,
> > 
> > a. Debian or Raspberry Pi OS bullseye updated to bookworm?

OK: I checked. That's the Gunnar Wolf "unofficial only because it 
includes non-free firmware" Bookworm from raspi.debian.net / cdimage.debian.org

> > 
> > b. Why testing/Bookworm specifically?
> > 
> cuz I gave up trying to make a realtime kernel work on raspios's 
> bullseye.
> 

As before, you're pretty much on your own with realtime kernels on RPi.
I'd still suggest using stable Debian - so Bullseye - because testing
has the habit of large numbers of changes and you could end up with
problems in the next >1 year until Bookworm is released. And this is
64 bit under Debian - so RaspiOS knowledge won't really help anyway.

> > c. Your machine or somebody else's. If somebody else's, I'd really
> > recommend sticking to Bullseye as more people can help.
> 

See above: you could end up being a long distance support provider. Been
there, done it, got the T-shirt :( Your choices are not necessarily my 
choices, however, and I need to learn to respect that.

> Will be someone else's about 3000 miles west of me. I have 2, 64G u-sd 
> cards I'm installing on but when they work in my machine, will be sent 
> back to him. In the process, I'm using my rpi4 as a test mule.
>  

Quite a distance and possibly a few TZ - not my favourite for tech
support if it goes wrong.

> > Q2. There probably _is_ a first user. If this is Raspberry Pi OS, the
> > first user will be user pi, with gid 1000
> 

OK - so Gunnar's images specifically do _not_ have a first user set up.
They do have a "login for root with no password" / allow ssh as root set up.
That's OK in limited circumstances - you might want to go back and reread
the rationale for Gunnar's decisions. if there's a root user with no password
and effectively autologin, you'll need to rethink that.

> I've looked at /edtc/passwd, pi does not exist in it.
> 

See above

> > Users will only be added to sudo if you do so. if you set up Debian,
> > and explicitly say you do not want a root user / set a blank password
> > then the first user _will_ be added to sudo as a special case.
> > 

See above: if there's a root account set, sudo is pretty much moot.

> > All the very best, as ever,
> 
> While this confirms its Debian's version, built this past Friday for 
> armhf, it does nothing to answer my first question above. The other end 
> user has a similar situation in that he is at least 40 miles from the 
> nearest ipv6 capable connection. To me, this is a pita, if the guys that 
> write the dhcp code would get out of their mothers basement and visit the 
> real world, they would check to see if there was a working dns at its 
> usual ipv6 address, AND REVERT to ipv4 automatically if its not found! 
> The existing situation leaves it up to us to figure out how to disable it 
> unconditionally for every install we do. ipv4 works just fine out here in 
> the puckerbrush, if we can get ipv6 the hell out of the way.
> 

Do you _know_ exactly how his router / DNS / ISP connection is set up _in
advance_ ? A lot of folk have an ISP cable connection where the router
does everything. You can set up the router to provide DNS, acquire hostnames
for each machine, fix IP leases locked to MAC address. 

if he doesn't want to connect a CNC mill to anything, what does an IP address
matter anyway.

> I'm not frustrated with you Andy. But I think Debian has some pretty 
> smart coders who seem to be wearing tunnelvision glasses and can't see 
> the bigger picture that exists 5 miles away from the isp's premises.
> 
> That's frustrating, and should be fixed. Looking at the acreage where 
> ipv6 IS available, I'd guess it is less than 1% of this whole country, 
> why shove this down our throat?
> 

That's maybe true for you, but not for everybody.
At least knowing that it's mostly Debian is a help.

You may well find https://raspi.debian.net/defaults-and-settings/ useful

All best, as ever,

Andy Cater
> > Andy Cater
> > 
> > 
> > 
> > .
> 
> 
> Cheers, Gene Heskett.
> -- 
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author, 1940)
> If we desire respect for the law, we must first make the law respectable.
>  - Louis D. Brandeis
> Genes Web page <http://geneslinuxbox.net:6309/gene>
> 
> 
> 


Reply to: