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

Re: quasi-SOLVED: Re: no network after jessie -> stretch



David Wright wrote on 06/24/2017 06:42 PM:

> 
>>> The error message above indicates, that you have network-manager
>>> installed and since stretch NetworkManager-wait-online.service is
>>> enabled by default (it wasn't in jessie).
> 
> this would suggest a cause. Do you need the network before your login
> prompt appears or not? If not, it looks like systemd needs telling that.
> I think this just came up in a contemporaneous thread here, but in a
> disk-mounting context rather than networking.
> 

Well, some remote disk filesystems get mounted at boot time -- so that
obviously needs a functioning network -- but I have no idea whether that
happens before the login prompt appears. It is definitely true that by the
time I am logged in (which is very fast, since I use i3), the network is up
and running.

I'll go look at the disk-mounting thread and see if there's something useful
there.

>>> Now, if you don't actually manage your interfaces with NetworkManager,
>>> NetworkManager-wait-online.service might run into a timeout (of 30s).
> 
> You originally wrote "about 30 useless seconds". It is worth stating
> whether that is a counted-down timeout period or just an estimate of
> a period you have wait. (I'm assuming you've done nothing to increase
> the flow of console messages when you boot.)

This is a counted-up period. I forget precisely what appears on the screen,
but it's a counter that counts upwards; and, strangely, it doesn't suggest
that there's a 30-second timeout (it's of the form "<nnn>s / No timeout", or
something like that, where the <nnn> increments each second). But when it hits
30 seconds, it does time out and the rest of the boot sequence occurs.

> 
> I have no idea whether the status of eth1 could have any bearing
> because I don't know where one makes ones wishes known to NM,
> not having used it.

Ditto. (At least, "not having used it knowingly".)

Nothing is plugged into eth1, and I have never told the system to configure
that interface, so one would hope that it would be completely ignored. But I
have no idea if that's actually what happens. The error message from NM is
singularly unhelpful :-(

  Doc

-- 
Web:  http://enginehousebooks.com/drevans

Attachment: signature.asc
Description: OpenPGP digital signature


Reply to: