Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Problem installing OS with netboot.xyz
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

Problem installing OS with netboot.xyz

Have a VPS where you need to set static IPv4 when you install an OS. Works fine for most ISOs but am unable to get netboot.xyz to install an OS.

I boot to netboot ISO and it will just go into a reboot loop until I select M for Manual. Asks for IP address: 123.456.789.012 Asks for Subnet Mask: 255.255.255.0. Asks for Gateway: 123.456.789.1. Asks for DNS: 8.8.8.8

At this point netboot starts up and displays a list of OSs. When I choose one it begins the install but apparently there is a problem with network so is unable to proceed.

Believe I need to enter the Network or Subnet info somewhere (123.456.789.0/24) but it's not giving me the option.

Any info or advice is appreciated.

Comments

  • Daniel15Daniel15 Veteran
    edited October 2021

    When I choose one it begins the install but apparently there is a problem with network so is unable to proceed.

    Which OS? On Debian you'll have to wait for DHCP to fail, then choose to manually configure the network, and enter all the details. This works for most hosts for me, but failed on ZappieHost because their gateway is outside the subnet (you may need to open a ticket and ask them for assistance in that case).

    It's even easier if you use grub-imageboot. I've commented about this in the past on this forum, and just wrote a page on the Netboot.xyz site about how to do it: https://netboot.xyz/docs/booting/grub. Note that on most KVM VPSes, you can just press a key on boot and select iPXE, no existing OS required.

    In any case, that's not really relevant to this thread I don't think? OP said they can boot netboot.xyz and start the installer of some OS, they just can't complete the install.

    Thanked by 2dosai JDMcPea
  • angstromangstrom Moderator
    edited October 2021

    @JDMcPea said: Any info or advice is appreciated.

    I've experienced the same issue with netboot.xyz before.

    My frank advice in this case would be to forget netboot.xyz and instead to try to get the particular ISO that you want to install. For example, if you want to install Debian, try to get a Debian ISO, and once you boot the ISO, go into "Expert install" and there's an option for static network settings.

    I imagine that netboot.xyz can be made to work with static network settings, but after several tries I gave up and went directly for the ISO that I wanted to install (Debian, in my case).

    Thanked by 1JDMcPea
  • @Daniel15 said:

    When I choose one it begins the install but apparently there is a problem with network so is unable to proceed.

    Which OS? On Debian you'll have to wait for DHCP to fail, then choose to manually configure the network, and enter all the details. This works for most hosts for me, but failed on ZappieHost because their gateway is outside the subnet (you may need to open a ticket and ask them for assistance in that case).

    How do you have a gateway outside the subnet (intentionally)? Are you sure the GW wasn't an accidental typo by provider? That's happened to me before.

    Thanked by 1JDMcPea
  • Daniel15Daniel15 Veteran
    edited October 2021

    @TimboJones said: How do you have a gateway outside the subnet (intentionally)? Are you sure the GW wasn't an accidental typo by provider?

    That's what I thought, but somehow it actually works? My guess is the host machine has multiple IP ranges, and one gateway for it to reach the internet, and the VMs on the host all share that same gateway. As long as all traffic is routed via the host via the virtual network adapter and the host knows how to reach the gateway (could even be a static route on the host), it probably works fine?

    A South African VPS I've got with ZappieHost has its IP in the 169.239.128.0/23 range, but the gateway is 129.232.177.186, and it works fine. A New Zealand VPS with the same provider has a gateway of 10.0.0.1 🤔 (although they did say they have NAT VPSes on the same node)

  • @Daniel15 said:

    @TimboJones said: How do you have a gateway outside the subnet (intentionally)? Are you sure the GW wasn't an accidental typo by provider?

    That's what I thought, but somehow it actually works? My guess is the host machine has multiple IP ranges, and one gateway for it to reach the internet, and the VMs on the host all share that same gateway. As long as all traffic is routed via the host via the virtual network adapter and the host knows how to reach the gateway (could even be a static route on the host), it probably works fine?

    No, it would be destination unreachable. It's invalid case.

    A South African VPS I've got with ZappieHost has its IP in the 169.239.128.0/23 range, but the gateway is 129.232.177.186, and it works fine. A New Zealand VPS with the same provider has a gateway of 10.0.0.1 🤔 (although they did say they have NAT VPSes on the same node)

    I call shenanigans. Still got the ZappieHost VM? Post the routing table. There'd need to be a static route for "You want 129.232.177.186? Send it to 169.239.128.1", which could just replace the wrong gw with the address used for static route.

  • Daniel15Daniel15 Veteran
    edited October 2021

    @TimboJones said: I call shenanigans. Still got the ZappieHost VM? Post the routing table.

    I also call shenanigans, but somehow it works. It also has a 255.255.255.255 netmask which is strange to me too. Their support told me that the Debian installer doesn't like it, so they temporarily gave me an address in the 129.232.177.x range just for installation, but then once I rebooted I changed the IP and it works fine.

    daniel@za02:~$ cat /etc/network/interfaces
    # This file describes the network interfaces available on your system
    # and how to activate them. For more information, see interfaces(5).
    
    source /etc/network/interfaces.d/*
    
    auto lo
    iface lo inet loopback
    
    allow-hotplug ens18
    iface ens18 inet static
            address 169.239.129.38/32
            gateway 129.232.177.186
    
    iface ens18 inet6 static
            address 2c0f:f530::d00:332/128
            gateway 2c0f:fce8:4000:1006::1
    
    daniel@za02:~$ ip r
    default via 129.232.177.186 dev ens18 onlink
    
    daniel@za02:~$ curl https://ipv4.icanhazip.com/
    169.239.129.38
    
    daniel@za02:~$ traceroute -4 google.com
    traceroute to google.com (172.217.170.78), 30 hops max, 60 byte packets
     1  129.232.177.186 (129.232.177.186)  0.065 ms  0.036 ms  0.032 ms
     2  249.165.232.129.reverse.xneelo.net (129.232.165.249)  2.021 ms  1.994 ms  1.973 ms
     3  core-access-switch1-v4001.jnb1.host-h.net (197.189.193.33)  10.363 ms  10.119 ms  10.333 ms
     4  core-router1.jnb1.host-h.net (197.189.193.2)  0.740 ms  0.707 ms core-router2.jnb1.host-h.net (197.189.193.3)  19.605 ms
     5  google.jinx.net.za (196.223.14.41)  1.631 ms  1.613 ms  1.597 ms
     6  * * *
     7  172.253.65.248 (172.253.65.248)  1.291 ms jnb02s04-in-f14.1e100.net (172.217.170.78)  3.107 ms 172.253.65.254 (172.253.65.254)  3.526 ms
    
    daniel@za02:~$ ip a
    1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
        link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
        inet 127.0.0.1/8 scope host lo
           valid_lft forever preferred_lft forever
        inet6 ::1/128 scope host
           valid_lft forever preferred_lft forever
    2: ens18: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
        link/ether -removed- brd ff:ff:ff:ff:ff:ff
        altname enp0s18
        inet 169.239.129.38/32 brd 169.239.129.38 scope global ens18
           valid_lft forever preferred_lft forever
        inet6 2c0f:f530::d00:332/128 scope global
           valid_lft forever preferred_lft forever
        inet6 fe80::58af:90ff:fecb:6b66/64 scope link
           valid_lft forever preferred_lft forever
    
    Thanked by 2yoursunny TimboJones
  • @Daniel15 said: default via 129.232.177.186 dev ens18 onlink

    onlink -> pretend that the nexthop is directly attached to this link, even if it does not match any interface prefix.

  • @JabJab said:

    @Daniel15 said: default via 129.232.177.186 dev ens18 onlink

    onlink -> pretend that the nexthop is directly attached to this link, even if it does not match any interface prefix.

    Interesting... I never manually set this; I just have gateway 129.232.177.186 in the network config. I wonder if Debian automatically does that?

  • @Daniel15 said: I wonder if Debian automatically does that?

    ifupdown (0.8.7) unstable; urgency=medium
    (...)
    * Use the onlink option when adding a gateway route. Closes: #378506

    -- Guus Sliepen <> Mon, 11 Jan 2016 22:58:44 +0100

    Thanked by 1Daniel15
  • Daniel15Daniel15 Veteran
    edited October 2021

    Cool thanks @JabJab; I learned something today :smiley:

  • Thanks for the responses. Have taken @angstrom advice and given up on netboot for this VPS, using another ISO host provides where you can just enter all the network info during setup and works no issues

  • @JabJab said:

    @Daniel15 said: default via 129.232.177.186 dev ens18 onlink

    onlink -> pretend that the nexthop is directly attached to this link, even if it does not match any interface prefix.

    Thanks, learned something new.

    Why would a provider do this? Or is this some openvz shit?

  • @TimboJones said: Or is this some openvz shit?

    My VPSes are both KVM.

    Thanked by 1TimboJones
  • @TimboJones said:

    @JabJab said:

    @Daniel15 said: default via 129.232.177.186 dev ens18 onlink

    onlink -> pretend that the nexthop is directly attached to this link, even if it does not match any interface prefix.

    Thanks, learned something new.

    Why would a provider do this? Or is this some openvz shit?

    Let's tag @Zappie and see - maybe they will answer :)

    Thanked by 1TimboJones
  • ZappieZappie Member, Host Rep, LIR

    Ha thanks for the tag.

    Cant speak for the issue that OP is having (to be honest I didnt fully read it sorry) but like @Daniel15 said, the debian installed (just the installer) doesnt like a fixed /32 subnet (255.255.255.255) unlike other installers like Centos for example which handle it just fine.

    But once in the OS (be it Ubuntu, Debian, CentOs, Arch and yes EVEN WINDOWS) its totally ok and legitimate to have a single IP with a 255.255.255.255 mask with with a gateway that is (obviously) outside that subnet. Iin your routing table you would see

    default via 1.2.3.4 dev eth0
    1.2.3.4 dev eth0 scope link
    

    1.2.3.4 being the gateway address. As far as I know (I could be wrong) is the exact same way that (for example) OpenVZ does networking where a single IP address is added and everything is added to the single gateway.

    Thanked by 1Daniel15
  • @Zappie said: single IP with a 255.255.255.255 mask

    The thing I don't quite understand about this is what the network and broadcast addresses are with a 255.255.255.255 netmask. Do they just... not exist? Network, broadcast, and a host sharing the same IP seems like it shouldn't work 🤔

  • ZappieZappie Member, Host Rep, LIR

    @Daniel15 said: The thing I don't quite understand about this is what the network and broadcast addresses are with a 255.255.255.255 netmask. Do they just... not exist? Network, broadcast, and a host sharing the same IP seems like it shouldn't work 🤔

    It just meant you cant do unicast with others on the network (since the network is only you) Most PPPoE ISPs and some VPNs do the exact same thing, dishing out a single address and a routed gateway to push all your traffic from and to

    Side note, this can also work with (for example) a /31 where you have 1.2.3.10/31 as the GW and 1.2.3.11/31 as the client and thats the full network.

    Thanked by 1Daniel15
  • ZappieZappie Member, Host Rep, LIR

    @JDMcPea said: Have a VPS where you need to set static IPv4 when you install an OS. Works fine for most ISOs but am unable to get netboot.xyz to install an OS.

    I boot to netboot ISO and it will just go into a reboot loop until I select M for Manual. Asks for IP address: 123.456.789.012 Asks for Subnet Mask: 255.255.255.0. Asks for Gateway: 123.456.789.1. Asks for DNS: 8.8.8.8

    In regards to this, it does sound very weird, if you were able to mount other (net)ISOs on your platform and give them your VPS static IP address/gateway and have them work all fine, I dont see why netboot specifically wont work 🤔 (dumb suggestions but maybe typo in the address?)

  • @Zappie said:
    Ha thanks for the tag.

    Cant speak for the issue that OP is having (to be honest I didnt fully read it sorry) but like @Daniel15 said, the debian installed (just the installer) doesnt like a fixed /32 subnet (255.255.255.255) unlike other installers like Centos for example which handle it just fine.

    But once in the OS (be it Ubuntu, Debian, CentOs, Arch and yes EVEN WINDOWS) its totally ok and legitimate to have a single IP with a 255.255.255.255 mask with with a gateway that is (obviously) outside that subnet. Iin your routing table you would see

    default via 1.2.3.4 dev eth0
    1.2.3.4 dev eth0 scope link
    

    1.2.3.4 being the gateway address. As far as I know (I could be wrong) is the exact same way that (for example) OpenVZ does networking where a single IP address is added and everything is added to the single gateway.

    But why do this and cause problems for installers instead of just using a regular routed network? There must be a payoff worth it.

  • yoursunnyyoursunny Member, IPv6 Advocate

    I encountered "IPv4 address and gateway not in same subnet" situation when installing Ubuntu 20 on @EvolutionHost Roubaix location (it's something similar to OVH failover IP).
    This is my notes on how I solved this problem.

    iPXE: ESC - 3 - CTRL+B
    set net0/ip 192.0.2.1
    set net0/netmask 255.255.255.0
    set net0/gateway 203.0.113.254
    set dns 1.1.1.1
    ifopen net0
    chain --autofree https://boot.netboot.xyz
    
    Ubuntu 20.04 Legacy
    network autoconfig will fail, select "do not configure" and proceed, and press ESC on "select a mirror" page, execute a shell:
    ip route add 203.0.113.254 dev ens3
    ip route add default via 203.0.113.254
    echo nameserver 1.1.1.1 >/etc/resolv.conf
    
Sign In or Register to comment.