2nd IP address on an interface

[Robert Moskowitz]

This is on a F20 arm system.  It SHOULD follow F20 rules...

first: cat /etc/udev/rules.d/70-persistent-net.rules

# Added by 'write_udev' for detected device 'eth0'.
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", 
ATTR{address}=="02:67:15:00:01:78", NAME="eth0"

First IP address is:

/etc/sysconfig/network-scripts/ifcfg-eth0

DEVICE="eth0"
BOOTPROTO=none
ONBOOT="yes"
TYPE="Ethernet"
NAME="System eth0"
MACADDR=02:67:15:00:01:63
MTU=1500
DNS1=208.83.67.188
GATEWAY="208.83.67.161"
IPADDR="208.83.67.163"
NETMASK="255.255.255.240"
HOSTNAME="miredo.htt-consult.com"
IPV6INIT="yes"
DNS2=2607:f4b8:3:3:9254:5400:0:188

this works fine.  But now for the second IP address I add:

# cat /etc/sysconfig/network-scripts/ifcfg-eth0:0
DEVICE="eth0:0"
BOOTPROTO=none
ONBOOT="yes"
TYPE="Ethernet"
NAME="System eth0"
MTU=1500
GATEWAY="208.83.67.161"
IPADDR="208.83.67.164"
NETMASK="255.255.255.240"

But no second IP address is shown:

]# ifconfig
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
         inet 208.83.67.163  netmask 255.255.255.240  broadcast 
208.83.67.175
         inet6 fe80::67:15ff:fe00:163  prefixlen 64  scopeid 0x20<link>
         ether 02:67:15:00:01:63  txqueuelen 1000  (Ethernet)
         RX packets 21  bytes 2556 (2.4 KiB)
         RX errors 0  dropped 12  overruns 0  frame 0
         TX packets 147  bytes 11726 (11.4 KiB)
         TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
         device interrupt 87  base 0x6000

lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 16436
         inet 127.0.0.1  netmask 255.0.0.0
         inet6 ::1  prefixlen 128  scopeid 0x10<host>
         loop  txqueuelen 0  (Local Loopback)
         RX packets 90  bytes 8076 (7.8 KiB)
         RX errors 0  dropped 0  overruns 0  frame 0
         TX packets 90  bytes 8076 (7.8 KiB)
         TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

So what is missing?  Do I need a 70-persistent-net.rules for eth0:0 ??

Or something else?

[Tom Horsley]

On Thu, 28 Aug 2014 16:16:55 -0400
Robert Moskowitz wrote:
> So what is missing?  Do I need a 70-persistent-net.rules for eth0:0 ??

These days, I'm pretty sure you are supposed to include
IPADDR2=, NETMASK2=, etc. in the one ifcfg-eth0 file
rather than creating a eth0:0 file (at least that worked
for me on centos 6.5 this week when setting up IP aliases).

[Robert Moskowitz]

On 08/28/2014 04:24 PM, Tom Horsley wrote:
> On Thu, 28 Aug 2014 16:16:55 -0400
> Robert Moskowitz wrote:
> > So what is missing?  Do I need a 70-persistent-net.rules for eth0:0 ??
>
> These days, I'm pretty sure you are supposed to include
> IPADDR2=, NETMASK2=, etc. in the one ifcfg-eth0 file
> rather than creating a eth0:0 file (at least that worked
> for me on centos 6.5 this week when setting up IP aliases).

I added the IPADDR2 and NETMASK2 and it did not add the second address.  
Even after a reboot.

[Ed Greshko]

On 08/29/14 06:11, Robert Moskowitz wrote:
> On 08/28/2014 04:24 PM, Tom Horsley wrote:
> > On Thu, 28 Aug 2014 16:16:55 -0400
> > Robert Moskowitz wrote:
> > > So what is missing?  Do I need a 70-persistent-net.rules for eth0:0 ??
> >
> > These days, I'm pretty sure you are supposed to include
> > IPADDR2=, NETMASK2=, etc. in the one ifcfg-eth0 file
> > rather than creating a eth0:0 file (at least that worked
> > for me on centos 6.5 this week when setting up IP aliases).
>
> I added the IPADDR2 and NETMASK2 and it did not add the second address.  Even after a reboot.

You may be getting confused by using "ifconfig".....

[egreshko@f20f network-scripts]$ cat ifcfg-p2p1
TYPE="Ethernet"
BOOTPROTO=none
DEFROUTE="yes"
IPV4_FAILURE_FATAL=yes
IPV6INIT=yes
IPV6_AUTOCONF=yes
IPV6_DEFROUTE="yes"
IPV6_FAILURE_FATAL="no"
NAME="p2p1"
UUID="5ce325cb-5048-48d7-bdc0-457f278fe1f1"
ONBOOT="yes"
DNS1=192.168.1.18
DOMAIN=greshko.com
HWADDR=08:00:27:B7:04:4A
IPADDR0=192.168.1.227
PREFIX0=24
GATEWAY0=192.168.1.1
IPADDR1=192.168.1.19
PREFIX1=24
IPV6_PEERDNS=yes
IPV6_PEERROUTES=yes

[egreshko@f20f network-scripts]$ ifconfig
enp0s3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.1.227  netmask 255.255.255.0  broadcast 192.168.1.255
        inet6 2001:470:d:6bd:a00:27ff:feb7:44a  prefixlen 64  scopeid 0x0<global>
        inet6 fe80::a00:27ff:feb7:44a  prefixlen 64  scopeid 0x20<link>
        ether 08:00:27:b7:04:4a  txqueuelen 1000  (Ethernet)
        RX packets 478  bytes 57719 (56.3 KiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 383  bytes 54641 (53.3 KiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

*BUT*

[egreshko@f20f network-scripts]$ ip addr
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 08:00:27:b7:04:4a brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.227/24 brd 192.168.1.255 scope global enp0s3
       valid_lft forever preferred_lft forever
    inet 192.168.1.19/24 brd 192.168.1.255 scope global secondary enp0s3
       valid_lft forever preferred_lft forever
    inet6 2001:470:d:6bd:a00:27ff:feb7:44a/64 scope global noprefixroute dynamic
       valid_lft 7200sec preferred_lft 600sec
    inet6 fe80::a00:27ff:feb7:44a/64 scope link
       valid_lft forever preferred_lft forever

And from a second system....

[egreshko@meimei ~]$ ping 192.168.1.19
PING 192.168.1.19 (192.168.1.19) 56(84) bytes of data.
64 bytes from 192.168.1.19: icmp_seq=1 ttl=64 time=0.337 ms
64 bytes from 192.168.1.19: icmp_seq=2 ttl=64 time=0.290 ms
64 bytes from 192.168.1.19: icmp_seq=3 ttl=64 time=0.269 ms

[Ed Greshko]

On 08/29/14 07:20, Ed Greshko wrote:
> And from a second system....
>
> [egreshko@meimei ~]$ ping 192.168.1.19
> PING 192.168.1.19 (192.168.1.19) 56(84) bytes of data.
> 64 bytes from 192.168.1.19: icmp_seq=1 ttl=64 time=0.337 ms
> 64 bytes from 192.168.1.19: icmp_seq=2 ttl=64 time=0.290 ms
> 64 bytes from 192.168.1.19: icmp_seq=3 ttl=64 time=0.269 ms

For completeness....

[egreshko@meimei ~]$ ssh 192.168.1.19
Last login: Fri Aug 29 07:15:07 2014 from meimei.greshko.com
[egreshko@f20f ~]$ exit
logout
Connection to 192.168.1.19 closed.

[egreshko@meimei ~]$ ssh 192.168.1.227
Last login: Fri Aug 29 07:31:22 2014 from meimei.greshko.com

[Tom Horsley]

On Fri, 29 Aug 2014 07:20:18 +0800
Ed Greshko wrote:
> You may be getting confused by using "ifconfig".....

Yea, I forgot that bit, my aliases didn't show up
in ifconfig, but I could ping them from another
computer on the same network.

SOLVED - Re: 2nd IP address on an interface

[Robert Moskowitz]

It is all ifconfigs fault   :)

I just assumed that in f20 it was still good enough until F21.....

# cat /etc/udev/rules.d/70-persistent-net.rules

# Added by 'write_udev' for detected device 'eth0'.
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", 
ATTR{address}=="02:67:15:00:01:78", NAME="eth0"

# cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE="eth0"
BOOTPROTO=none
ONBOOT="yes"
TYPE="Ethernet"
NAME="System eth0"
MACADDR=02:67:15:00:01:63
MTU=1500
DNS1=208.83.67.148
GATEWAY="208.83.67.161"
IPADDR="208.83.67.163"
IPADDR2="208.83.67.164"
NETMASK="255.255.255.240"
NETMASK2="255.255.255.240"
HOSTNAME="miredo.htt-consult.com"
IPV6INIT="yes"
DNS2=2607:f4b8:3:0:9254:5400:0:148

# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
group default
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
     inet 127.0.0.1/8 scope host lo
     inet6 ::1/128 scope host
        valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast 
state UP group default qlen 1000
     link/ether 02:67:15:00:01:63 brd ff:ff:ff:ff:ff:ff
     inet 208.83.67.163/28 brd 208.83.67.175 scope global eth0
     inet 208.83.67.164/28 brd 208.83.67.175 scope global secondary eth0
     inet6 fe80::67:15ff:fe00:163/64 scope link
        valid_lft forever preferred_lft forever

On 08/28/2014 07:37 PM, Tom Horsley wrote:
> On Fri, 29 Aug 2014 07:20:18 +0800
> Ed Greshko wrote:
> > You may be getting confused by using "ifconfig".....
>
> Yea, I forgot that bit, my aliases didn't show up
> in ifconfig, but I could ping them from another
> computer on the same network.

[Robert Moskowitz]

On 08/28/2014 07:20 PM, Ed Greshko wrote:
> On 08/29/14 06:11, Robert Moskowitz wrote:
> > On 08/28/2014 04:24 PM, Tom Horsley wrote:
> > > On Thu, 28 Aug 2014 16:16:55 -0400
> > > Robert Moskowitz wrote:
> > > > So what is missing?  Do I need a 70-persistent-net.rules for eth0:0 ??
> > >
> > > These days, I'm pretty sure you are supposed to include
> > > IPADDR2=, NETMASK2=, etc. in the one ifcfg-eth0 file
> > > rather than creating a eth0:0 file (at least that worked
> > > for me on centos 6.5 this week when setting up IP aliases).
> >
> > I added the IPADDR2 and NETMASK2 and it did not add the second address.  Even after a reboot.
>
> You may be getting confused by using "ifconfig".....
>
> [egreshko@f20f network-scripts]$ cat ifcfg-p2p1
> TYPE="Ethernet"
> BOOTPROTO=none
> DEFROUTE="yes"
> IPV4_FAILURE_FATAL=yes
> IPV6INIT=yes
> IPV6_AUTOCONF=yes
> IPV6_DEFROUTE="yes"
> IPV6_FAILURE_FATAL="no"
> NAME="p2p1"
> UUID="5ce325cb-5048-48d7-bdc0-457f278fe1f1"
> ONBOOT="yes"
> DNS1=192.168.1.18
> DOMAIN=greshko.com
> HWADDR=08:00:27:B7:04:4A
> IPADDR0=192.168.1.227
> PREFIX0=24
> GATEWAY0=192.168.1.1
> IPADDR1=192.168.1.19
> PREFIX1=24
> IPV6_PEERDNS=yes
> IPV6_PEERROUTES=yes
>
> [egreshko@f20f network-scripts]$ ifconfig
> enp0s3: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
>          inet 192.168.1.227  netmask 255.255.255.0  broadcast 192.168.1.255
>          inet6 2001:470:d:6bd:a00:27ff:feb7:44a  prefixlen 64  scopeid 0x0<global>
>          inet6 fe80::a00:27ff:feb7:44a  prefixlen 64  scopeid 0x20<link>
>          ether 08:00:27:b7:04:4a  txqueuelen 1000  (Ethernet)
>          RX packets 478  bytes 57719 (56.3 KiB)
>          RX errors 0  dropped 0  overruns 0  frame 0
>          TX packets 383  bytes 54641 (53.3 KiB)
>          TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>
> *BUT*
>
> [egreshko@f20f network-scripts]$ ip addr
> 2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
>      link/ether 08:00:27:b7:04:4a brd ff:ff:ff:ff:ff:ff
>      inet 192.168.1.227/24 brd 192.168.1.255 scope global enp0s3
>         valid_lft forever preferred_lft forever
>      inet 192.168.1.19/24 brd 192.168.1.255 scope global secondary enp0s3
>         valid_lft forever preferred_lft forever
>      inet6 2001:470:d:6bd:a00:27ff:feb7:44a/64 scope global noprefixroute dynamic
>         valid_lft 7200sec preferred_lft 600sec
>      inet6 fe80::a00:27ff:feb7:44a/64 scope link
>         valid_lft forever preferred_lft forever
>
> And from a second system....
>
> [egreshko@meimei ~]$ ping 192.168.1.19
> PING 192.168.1.19 (192.168.1.19) 56(84) bytes of data.
> 64 bytes from 192.168.1.19: icmp_seq=1 ttl=64 time=0.337 ms
> 64 bytes from 192.168.1.19: icmp_seq=2 ttl=64 time=0.290 ms
> 64 bytes from 192.168.1.19: icmp_seq=3 ttl=64 time=0.269 ms

# ip addr show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN 
group default
     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
     inet 127.0.0.1/8 scope host lo
     inet6 ::1/128 scope host
        valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast 
state UNKNOWN group default qlen 1000
     link/ether 02:56:02:01:f3:b9 brd ff:ff:ff:ff:ff:ff
     inet 208.83.67.163/28 brd 208.83.67.175 scope global eth0
     inet 208.83.67.164/28 brd 208.83.67.175 scope global secondary eth0
     inet6 fe80::56:2ff:fe01:f3b9/64 scope link
        valid_lft forever preferred_lft forever

you are right.  ifconfig USE to be able to do this, but now it is just a 
shell?  over ip and so...

So now to undo a lot of changes and see if I can get back to everything 
working.

[James Hogarth]

On 29 Aug 2014 01:32, "Robert Moskowitz"  wrote:
> you are right.  ifconfig USE to be able to do this, but now it is just a

shell?  over ip and so...
>

Your memory isn't quite accurate... ifconfig has never handled multiple IP
addresses on an interface with aliases being used to create separate
virtual interfaces instead.

It is also not a shell over ip but rather a completely different package
with a completely different development history...

This is just yet another example of why net-tools has been deprecated a
long time and that people really need to break their muscle memory and use
iproute2 ...

Is net-tools still in the base fedora install? Probably long since time to
pull it if so... I know el7 doesn't include it in base or core package
groups anymore at least..

[Robert Moskowitz]

On 08/29/2014 01:48 AM, James Hogarth wrote:
> On 29 Aug 2014 01:32, "Robert Moskowitz" <[e-mail] 
> > wrote:
> > you are right.  ifconfig USE to be able to do this, but now it is
>
> just a shell?  over ip and so...
> >
>
> Your memory isn't quite accurate... ifconfig has never handled 
> multiple IP addresses on an interface with aliases being used to 
> create separate virtual interfaces instead.

Well recently I was going by:

http://www.tecmint.com/create-multiple-ip-addresses-to-one-single-network-in...

Which show ifconfig listing all the ethn:m interfaces.  Also I use to do 
a lot of vlaning when I ran radvd on linux before my firewall supported 
it.  Now maybe that was back in Centos 4 days.
> It is also not a shell over ip but rather a completely different 
> package with a completely different development history...
>
> This is just yet another example of why net-tools has been deprecated 
> a long time and that people really need to break their muscle memory 
> and use iproute2 ...

Working on it.  Really.
> Is net-tools still in the base fedora install? Probably long since 
> time to pull it if so... I know el7 doesn't include it in base or core 
> package groups anymore at least..

[Tom H]

On Thu, Aug 28, 2014 at 8:32 PM, Robert Moskowitz  wrote:
> # ip addr show
> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN group
> default
>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>     inet 127.0.0.1/8 scope host lo
>     inet6 ::1/128 scope host
>        valid_lft forever preferred_lft forever
> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state
> UNKNOWN group default qlen 1000
>     link/ether 02:56:02:01:f3:b9 brd ff:ff:ff:ff:ff:ff
>     inet 208.83.67.163/28 brd 208.83.67.175 scope global eth0
>     inet 208.83.67.164/28 brd 208.83.67.175 scope global secondary eth0
>     inet6 fe80::56:2ff:fe01:f3b9/64 scope link
>        valid_lft forever preferred_lft forever
>
> you are right. ifconfig USE to be able to do this, but now it is just a
> shell? over ip and so...
>
> So now to undo a lot of changes and see if I can get back to everything
> working.

ip (and ss, bridge, tc, and other iproute binaries) and ifconfig (and
route, netstat, brctl, ... and other net-tools and bridge-utils
binaries) use different kernel interfaces.

AIUI, the ifconfig interface is deprecated but kept for backward compatibility.

[Tom H]

On Thu, Aug 28, 2014 at 6:11 PM, Robert Moskowitz  wrote:
> On 08/28/2014 04:24 PM, Tom Horsley wrote:
> > These days, I'm pretty sure you are supposed to include
> > IPADDR2=, NETMASK2=, etc. in the one ifcfg-eth0 file
> > rather than creating a eth0:0 file (at least that worked
> > for me on centos 6.5 this week when setting up IP aliases).
>
> I added the IPADDR2 and NETMASK2 and it did not add the second address.
> Even after a reboot.

If you use IPADDR2, you have to display ip addresses with "ip a"
because "ifconfig" won't see the second address.

[Lars E. Pettersson]

On 08/28/14 22:16, Robert Moskowitz wrote:
> # cat /etc/sysconfig/network-scripts/ifcfg-eth0:0
> DEVICE="eth0:0"
> BOOTPROTO=none
> ONBOOT="yes"
> TYPE="Ethernet"
> NAME="System eth0"
> MTU=1500
> GATEWAY="208.83.67.161"
> IPADDR="208.83.67.164"
> NETMASK="255.255.255.240"

I think you need to add
ONPARENT=yes
to make it start when its parent does.

That is at least the major difference I see comparing to my setup.

Lars

[Robert Moskowitz]

On 08/28/2014 04:35 PM, Lars E. Pettersson wrote:
> On 08/28/14 22:16, Robert Moskowitz wrote:
> > # cat /etc/sysconfig/network-scripts/ifcfg-eth0:0
> > DEVICE="eth0:0"
> > BOOTPROTO=none
> > ONBOOT="yes"
> > TYPE="Ethernet"
> > NAME="System eth0"
> > MTU=1500
> > GATEWAY="208.83.67.161"
> > IPADDR="208.83.67.164"
> > NETMASK="255.255.255.240"
>
> I think you need to add
> ONPARENT=yes
> to make it start when its parent does.
>
> That is at least the major difference I see comparing to my setup.

Add this to ifcfg-eth0:0 and no change.

:(

Got to keep on trying.

[Lars E. Pettersson]

> 29 aug 2014 kl. 00:12 skrev Robert Moskowitz 
> Add this to ifcfg-eth0:0 and no change.
>
> :(

Ok. Another difference I see is that my interface is named :1, i.e. in my case wan:1 Not sure if that makes any difference though, was years since I set this up :)

Lars

[Robert Moskowitz]

On 08/28/2014 06:23 PM, Lars E. Pettersson wrote:
> > 29 aug 2014 kl. 00:12 skrev Robert Moskowitz 
> > Add this to ifcfg-eth0:0 and no change.
> >
> > :(
>
> Ok. Another difference I see is that my interface is named :1, i.e. in my case wan:1 Not sure if that makes any difference though, was years since I set this up :)

I took out the macaddr changes I was making, trying to go back to as 
much 'base' stuff as I could, and it still is not giving me a 2nd ip addr.

:(

[Andre Speelmans]

On Thu, Aug 28, 2014 at 10:35 PM, Lars E. Pettersson  wrote:
> On 08/28/14 22:16, Robert Moskowitz wrote:
> > # cat /etc/sysconfig/network-scripts/ifcfg-eth0:0
> > [snipped]
> > ONBOOT="yes"
> I think you need to add
> ONPARENT=yes
> to make it start when its parent does.

ONPARENT=yes is indeed the parameter instead of ONBOOT for aliases. It
would not matter here, as "yes" is the default value.
As far as I am aware the suggestion from Tom Horsley with things like
IPADDR2, NETMASK2 in the file of the parent is cuurently the preferred
method. (Not my preferred method, but then perhaps I am just stuck
with old stuff that I know.) :)

[Robert Moskowitz]

On 08/29/2014 01:12 AM, Andre Speelmans wrote:
> On Thu, Aug 28, 2014 at 10:35 PM, Lars E. Pettersson  wrote:
> > On 08/28/14 22:16, Robert Moskowitz wrote:
> > > # cat /etc/sysconfig/network-scripts/ifcfg-eth0:0
> > > [snipped]
> > > ONBOOT="yes"
> >
> > I think you need to add
> > ONPARENT=yes
> > to make it start when its parent does.
>
> ONPARENT=yes is indeed the parameter instead of ONBOOT for aliases. It
> would not matter here, as "yes" is the default value.
> As far as I am aware the suggestion from Tom Horsley with things like
> IPADDR2, NETMASK2 in the file of the parent is cuurently the preferred
> method. (Not my preferred method, but then perhaps I am just stuck
> with old stuff that I know.) :)

I have been digging into real old configs from systems long gone (like 
back in '08).  I used the ifcfg-ethn.m for vlaning by including 
VLAN=yes, and then 'm' was the vlan number.  And the time back then that 
I ran miredo I used the IPADDR2= method (miredo needs 2 IPv4 addresses).