[gentoo-user] Dhclient and buggy ISP

2018-07-20 Thread Matti Nykyri
Hey

I have a following with Telia ISP and their dhcp-server... I'm using 
dhclient to get dynamic IP from ISP and also to update DDNS. When 
dhclient does DHCPDISCOVER the ISP will issue a new lease for 43200 sec 
after a normal DHCPOFFER, DHCPREQUEST and DHCPACK cycle; and the 
interface gets bound to the given address. Now when the dhclient does a 
subsequent DHCPREQUEST, server will issue DHCP with a declining 
lease-time ending at the same time as original 43200s lease. By 
releasing the lease and doing a new DHCPDISCOVER the lease time is reset 
to 43200s. The ISP how ever keeps track of the MAC and will always issue 
the same IP.

When the lease-time of 43200 runs out the default gw stops responding 
and the internet connection dies. The only way to then get it to work is 
to manually start a new discover cycle (ifdown - ifup).

I would like to configure the interface so, that there would not be an 
interuption in network traffic! Because the IP doesn't change, I would 
like to configure the WAN interface to stay up with the IP all the time 
and get dhclient to do DHCPDISCOVER to the ISP to update ISP-firewall to 
let the traffic through for the next 43200 sec. If the IP changes for 
some reason, I would like the dhclient to update the WAN interface.

Would there be a ready way in Gentoo to achieve this? Or anyway :'D

Ps. I hate my ISP :(
Pps. I have other problems too...

-- 
-Matti



Re: [gentoo-user] Re: scanner problem : latest

2018-07-20 Thread Philip Webb
180719 Ian Zimmerman wrote:
> On 2018-07-19 05:29, Philip Webb wrote:
>> The Mint version seems to find a file in /usr/share/iscan-data ,
>> which the others don't find, but that file is present in my Gentoo system.
> I have not looked at the strace log, but this smacks of permissions
> and possibly Selinux, if you've let that beast loose.

I checked the permissions for  iscan-data  & they seem ok.
No, I've never tried to use Selinux.

However, I've just escaped fr another beast, ie Linux Mint 19 :
DON'T try to install it alongside Gentoo !!

This beast leads you along to a point where you've told it where to install
-- a partition on my HDD (sdb5) -- ,
you click 'proceed' & it chugs along nicely for some time,
then it says it's trying to install a bootloader
without asking whether you want it to or where to do it.
Next, it says it's failed to install Grub -- which I've never used -- ,
offers several ugly-looking options & freezes.

All you can do is a hard power-off & when you restart,
instead of the regular Lilo box of options, there's only a Grub prompt.
Disaster looks you in the face !

SystemRescue to the rescue : Gentoo  /boot  is unaltered,
so copy  lilo.conf  into the SR  /etc  dir, edit it to use SR paths,
mount the Gentoo  /boot , run 'lilo' & the MBR is restored to what it sb.
Reboot & the familiar Lilo screen appears !

I wanted to see whether the latest Mint found my scanner,
which would suggest that the problem doesn't involve the kernel, at least.
I'm not trying again !

On inspection, Mint 19 does seem to have installed itself on  /dev/sdb5 ,
so I can copy the vital files from its  /boot  to my own Gentoo  /boot ,
edit  lilo.conf  accordingly & see if it will start : not today !

Next up -- too late for today -- is to see if I can get Nvidia working
with an earlier kernel : at present it isn't.
Then I can find out if Xsane can find my scanner with another kernel.

Can anyone take a look at my Strace output ( 3  files),
which is at  http://chass.utoronto.ca/~purslow/test ?
That would be a real help.

-- 
,,
SUPPORT ___//___,   Philip Webb
ELECTRIC   /] [] [] [] [] []|   Cities Centre, University of Toronto
TRANSIT`-O--O---'   purslowatchassdotutorontodotca




Re: [gentoo-user] Re: scanner problem : PS

2018-07-20 Thread Philip Webb
180720 Philip Webb wrote:
> On inspection, Mint 19 does seem to have installed itself on  /dev/sdb5 ,
> so I can copy the vital files from its  /boot  to my own Gentoo  /boot ,
> edit  lilo.conf  accordingly & see if it will start : not today !

I have successfully booted into Mint 19
-- it installed fully before it did its dirty MBR trick -- ,
so I can see what happens with Xsane tomorrow.

-- 
,,
SUPPORT ___//___,   Philip Webb
ELECTRIC   /] [] [] [] [] []|   Cities Centre, University of Toronto
TRANSIT`-O--O---'   purslowatchassdotutorontodotca




[gentoo-user] Computer keeps crashing during boot, need dmesg

2018-07-20 Thread Andrew Lowe
Hi all,
I'm having all sorts of trouble getting a new computer to boot. I have
two NVME thingies in it with one containing Win10 and the other,
eventually, Gentoo. I use Refind to control the booting with the machine
using UEFI.

The graphical front end to Refind boots fine and allows me to in turn
boot Windows. It also spots and then allows me to boot a copy of
SysrescueCD that I left plugged in. No joy with Linux. I get up to where
the kernel appears to be doing something with sata and then it crashes.
Now herein lies the problem.

Is there a way to get dmesg logged to a file so that I can see what's
actually failing? I need to be able to get back into the machine with
SysrescueCD, which is no problem, and then have a look at the dmesg
output to see what's going wrong.

Does anyone have any idea as to how to do this, or another way of doing
the debugging I'm trying to do? The stuff I've found on the web assumes
that some of the loggers are working and can do stuff, I'm not getting
to the stage where they have started up yet.

I'm in the process of transferring from a non EFI hard disk setup to
EFI & nvme. The hard disk based system boots with no problems. Assuming
that the hard disk system and the nvme system will give nearly the same
dmesg, looking at the hard disk dmesg, I see that if I go to where I
think is the same posi as the crash on the nvme, is where it attempts to
mount root:

VFS: Mounted root (jfs filesystem) readonly on device 8:19.

If I have managed to do something wrong with respect to this mounting,
would it cause the crash or just complain and freeze?

Any thoughts on this would be greatly appreciated,

Andrew




Re: [gentoo-user] Computer keeps crashing during boot, need dmesg

2018-07-20 Thread Andrew Lowe
On 20/07/18 21:26, Andrew Lowe wrote:
> Hi all,
>   I'm having all sorts of trouble getting a new computer to boot. I have

[snip]

> 
> VFS: Mounted root (jfs filesystem) readonly on device 8:19.
>

Forgot to add, in the new nvme config, jfs is a built in driver. Has
anyone heard of problems with jfs and nvme? I have a recollection
reading someone, somewhere saying that something needed to be set in
fstab if using jfs as root.

Andrew



Re: [gentoo-user] Re: scanner problem : latest

2018-07-20 Thread Mick
On Friday, 20 July 2018 14:04:26 BST Philip Webb wrote:
> 180719 Ian Zimmerman wrote:
> > On 2018-07-19 05:29, Philip Webb wrote:
> >> The Mint version seems to find a file in /usr/share/iscan-data ,
> >> which the others don't find, but that file is present in my Gentoo
> >> system.
> > 
> > I have not looked at the strace log, but this smacks of permissions
> > and possibly Selinux, if you've let that beast loose.
> 
> I checked the permissions for  iscan-data  & they seem ok.
> No, I've never tried to use Selinux.
> 
> However, I've just escaped fr another beast, ie Linux Mint 19 :
> DON'T try to install it alongside Gentoo !!
> 
> This beast leads you along to a point where you've told it where to install
> -- a partition on my HDD (sdb5) -- ,
> you click 'proceed' & it chugs along nicely for some time,
> then it says it's trying to install a bootloader
> without asking whether you want it to or where to do it.
> Next, it says it's failed to install Grub -- which I've never used -- ,
> offers several ugly-looking options & freezes.
> 
> All you can do is a hard power-off & when you restart,
> instead of the regular Lilo box of options, there's only a Grub prompt.
> Disaster looks you in the face !

Many an installation were hosed lately because of a GRUB problem in Mint and 
potentially *buntu.  Testing before releasing was not as thorough as it should 
have been:

https://blog.linuxmint.com/?p=3620

-- 
Regards,
Mick

signature.asc
Description: This is a digitally signed message part.


Re: [gentoo-user] Dhclient and buggy ISP

2018-07-20 Thread Mick
On Friday, 20 July 2018 08:58:53 BST Matti Nykyri wrote:
> Hey
> 
> I have a following with Telia ISP and their dhcp-server... I'm using
> dhclient to get dynamic IP from ISP and also to update DDNS. When
> dhclient does DHCPDISCOVER the ISP will issue a new lease for 43200 sec
> after a normal DHCPOFFER, DHCPREQUEST and DHCPACK cycle; and the
> interface gets bound to the given address. Now when the dhclient does a
> subsequent DHCPREQUEST, server will issue DHCP with a declining
> lease-time ending at the same time as original 43200s lease. By
> releasing the lease and doing a new DHCPDISCOVER the lease time is reset
> to 43200s. The ISP how ever keeps track of the MAC and will always issue
> the same IP.
> 
> When the lease-time of 43200 runs out the default gw stops responding
> and the internet connection dies. The only way to then get it to work is
> to manually start a new discover cycle (ifdown - ifup).
> 
> I would like to configure the interface so, that there would not be an
> interuption in network traffic! Because the IP doesn't change, I would
> like to configure the WAN interface to stay up with the IP all the time
> and get dhclient to do DHCPDISCOVER to the ISP to update ISP-firewall to
> let the traffic through for the next 43200 sec. If the IP changes for
> some reason, I would like the dhclient to update the WAN interface.
> 
> Would there be a ready way in Gentoo to achieve this? Or anyway :'D
> 
> Ps. I hate my ISP :(
> Pps. I have other problems too...


I am not familiar with dhclient, although whenever I used it in various 
distros it just worked.  Have you tried using dhcpcd instead?
-- 
Regards,
Mick

signature.asc
Description: This is a digitally signed message part.