Advanced KaVLAN

See also: Hardware | KaVLAN | Network reconfiguration tutorial | Advanced KaVLAN | Advanced IPv6 | Reconfigurable firewall | Network emulation

	Note
	This page is actively maintained by the Grid'5000 team. If you encounter problems, please report them (see the Support page). Additionally, as it is a wiki page, you are free to make minor corrections yourself if needed. If you would like to suggest a more fundamental change, please contact the Grid'5000 team.

Overview

The goal of Kavlan is to provide network isolation for Grid'5000 users. KaVLAN allow users to manage VLAN on their Grid'5000 nodes. The benefits is complete level 2 isolation. It can be used together with OAR and Kadeploy to do some experimentations on the platform.

The first step is to read the KaVLAN introduction to understand what kind of VLANs you can configure.

If you want a more concrete example of what you can do with VLANs on Grid'5000, you can go through the Network_reconfiguration_tutorial.

Reserve VLANs and deploy nodes inside

In the first part of the tutorial, we will use kadeploy and kavlan together on a single site, with a routed vlan (we could also use a local vlan).

Once connected on a frontend, in order to obtain nodes and a VLAN you must reserve a kavlan resource with oarsub. There are 3 kinds of resources: kavlan, kavlan-local, kavlan-global. Here, we will use 3 nodes and a routed VLAN, let's say in Sophia on cluster Suno:

frontend:

oarsub -t deploy -l {"type='kavlan'"}/vlan=1,{"cluster='suno'"}/nodes=3 -I

A shell is now opened on the frontend (like any regular deploy job) You can get the id of your VLAN using the kavlan command:

frontend:

kavlan -V

If you run this command outside the shell started by OAR for your reservation, you must add the oar JOBID.

frontend:

kavlan -V -j JOBID

You should get an integer in the <4-9> range for this routed VLAN (the range for local vlan is <1-3>, and there is one global VLAN per OAR server, i.e. one per site).

For our example, let's say we got suno-2, suno-30 and suno-31, and kavlan #4.

You can get all the options of the command using --help:

# kavlan --help
Usage: kavlan [options]
Specific options:
    -i, --vlan-id N                  set VLAN ID (integer or DEFAULT)
    -C, --ca-cert CA                 CA certificate
    -c, --client-cert CERT           client certificate
    -k, --client-key KEY             client key
    -l, --get-nodelist               Show nodenames in the given vlan
    -e, --enable-dhcp                Start DHCP server
    -d, --disable-dhcp               Stop DHCP server
    -V, --show-vlan-id               Show vlan id of job (needs -j JOBID)
    -g, --get-vlan                   Show vlan of nodes
    -s, --set-vlan                   Set vlan of nodes
    -j, --oar-jobid JOBID            OAR job id
    -m, --machine NODE               set nodename (several -m are OK)
    -f, --filename NODEFILE          read nodes from a file
    -u, --user USERNAME              username
    -v, --[no-]verbose               Run verbosely
    -q, --[no-]quiet                 Run quietly
        --[no-]debug                 Run with debug output
    -h, --help                       Show this message
        --version                    Show version

Once you have a kavlan reservation running, you can put your nodes in your VLAN (and back into the default VLAN) at anytime during the lifetime of your job; we will not use this for now.

Instead we will change the VLAN with kadeploy directly. The next step is to deploy the nodes with an environment image, for instance debian10-x64-big.

Enable the dhcp server of the VLAN

Before deploying, if you don't install your own DHCP server, you should start the default DHCP server of the VLAN. Do this with the kavlan command (add -j JOBID if needed) :

frontend:

kavlan -e

(You can disable the DHCP server with kavlan -d)

Deploy nodes and change VLAN in one step

frontend:

kadeploy3 -f $OAR_NODEFILE -k -e debian10-x64-big --vlan `kavlan -V`

Once the deployment is done, you will be able to connect on your nodes. They are now inside the VLAN, therefore they are not reachable with their default IP:

frontend:

ping suno-30 -c1

PING suno-30.sophia.grid5000.fr (172.16.130.30) 56(84) bytes of data.
From fsophia.sophia.grid5000.fr (172.16.143.106) icmp_seq=1 Destination Host Unreachable

--- suno-30.sophia.grid5000.fr ping statistics ---
1 packets transmitted, 0 received, +1 errors, 100% packet loss, time 0ms

You can get the list of new hostnames of your nodes in the VLAN with kavlan -l. For the next part of this tutorial, let's create a nodefile and copy it on the first node:

frontend:

kavlan -l | tee nodefile

suno-2-kavlan-4.sophia.grid5000.fr
suno-30-kavlan-4.sophia.grid5000.fr
suno-31-kavlan-4.sophia.grid5000.fr

frontend:

scp nodefile root@`head -1 < nodefile`:/tmp

The password for user root on Grid'5000 environments is "grid5000".

You can see that you can ping these new hostnames:

frontend:

ping suno-30-kavlan-4 -c1

PING suno-30-eth0-kavlan-4.sophia.grid5000.fr (10.32.3.30) 56(84) bytes of data.
64 bytes from suno-30-eth0-kavlan-4.sophia.grid5000.fr (10.32.3.30): icmp_seq=1 ttl=63 time=0.151 ms

--- suno-30-eth0-kavlan-4.sophia.grid5000.fr ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.151/0.151/0.151/0.000 ms

Setup a DHCP server on your nodes

Configure DHCP

If you need to run your own DHCP server (for example if you want to run a cluster distribution inside kavlan or test kadeploy ), you can use a script to generate the configuration file:

Then, go back the the frontend, and download the script that will generate your dhcp configuration:

Create this file (gen_dhcpd_conf.rb) on the frontend :

#!/usr/bin/ruby

# Author: Nicolas Niclausse
# Copyright 2010-2011: INRIA

# script specific to grid5000:
# generate dhcpd config files for kavlan

require 'rubygems'
require 'restfully' # gem install restfully --source http://gemcutter.org
require 'ip' # gem install ruby-ip
require 'getoptlong'
require 'optparse'
require 'ostruct'

headers = "ddns-update-style none;
option space pxelinux;
option pxelinux.magic      code 208 = string;
option pxelinux.configfile code 209 = text;
option pxelinux.pathprefix code 210 = text;
option pxelinux.reboottime code 211 = unsigned integer 32;
option vendorinfo          code 43  = string;
"

conf = File.expand_path('~/.restfully/api.grid5000.fr.yaml')
options = if FileTest.exists?(conf) then YAML.load_file(conf) else {} end
options[:base_uri] = 'https://api.grid5000.fr/stable/grid5000'

def parseopts(args)
  options = OpenStruct.new
  options.debug = false
  options.verbose = false
  options.quiet = false
  options.nodes = []
  opts = OptionParser.new do |opts|
    opts.banner = "Usage: gen_dhcpd_conf.rb [options]"
    opts.separator ""
    opts.separator "Specific options:"
    opts.on("-s","--site SITE",  "generate only DHCP conf for site SITE") do |site|
      options.site = site
    end
    opts.on("-i","--vlan-id N", Integer , "generate only DHCP conf for vlan N") do |vlan|
      options.vlan = vlan
    end
    opts.on("-q", "--[no-]quiet", "Run quietly") do |q|
      options.quiet = q
    end
    opts.on("-v", "--[no-]verbose", "Run verbosely") do |v|
      options.verbose = v
    end
    opts.on_tail("-h", "--help", "Show this message") do
      puts opts
      exit
    end
  end
  opts.parse!(args)
  options
end

$opts = parseopts(ARGV)

Restfully::Session.new(options) do |root, session|
  options = {:query => {:version => root['version']}}
  root.sites(options).each do |site|
    mysite=site['uid']
    next if not $opts.site.nil? and mysite != $opts.site
    # optionaly, read mac address from external yaml file
    ref = if FileTest.exists?(mysite+".yaml") then
            YAML.load_file(mysite+".yaml")
          else
            puts mysite +": no yaml file for macs" unless $opts.quiet
            {}
          end
    if $opts.vlan.nil? then
      vlans = (1..9).to_a
      # try to guess global vlan assigned to current site
      (10..21).each do |gvlan|
        begin
          IPSocket::getaddress("gw-kavlan-"+gvlan.to_s+"."+mysite+".grid5000.fr")
          puts "global vlan found for site %s: " % mysite unless $opts.quiet
          vlans.push(gvlan)
        rescue
          next
        end
      end
    else
      vlans = [$opts.vlan]
    end
    vlans.each do |vlan|
      filename = "dhcpd-kavlan-"+vlan.to_s+"-"+mysite+".conf"
      open(filename, 'w') do |f|
        puts "generating "+filename unless $opts.quiet
        f.puts headers
        begin
          gateway = IPSocket::getaddress("gw-kavlan-"+vlan.to_s+"."+mysite+".grid5000.fr")
        rescue
          puts "WARN: Get address error: probably no kavlan DNS setup for site " + mysite + " , skip" if $opts.verbose;
          next
        end
        # /20 for local vlans (1..3) and /18 for routed vlan (4..9)
        if vlan < 4
          ip = IP.new(gateway+"/20")
          ns = gateway
          ntp = gateway
          tftp = gateway
        else
          ip = IP.new(gateway+"/18")
          ntp = IPSocket::getaddress("ntp."+mysite+".grid5000.fr")
          ns = IPSocket::getaddress("dns."+mysite+".grid5000.fr")
          tftp = IPSocket::getaddress("kadeploy-server."+mysite+".grid5000.fr")
        end
        netmask = ip.netmask.to_addr
        broadcast = ip.broadcast.to_addr
        network = ip.network.to_addr
        f.puts "subnet %s netmask %s {" %  [network , netmask]
        f.puts "    default-lease-time 86400;
    max-lease-time 604800;"
        #f.puts "    option domain-name \"%s.grid5000.fr\"; " % mysite
        f.puts "    option domain-name-servers %s;" % ns
        f.puts "    option ntp-servers %s; " % ntp
        f.puts "    option routers %s;" % gateway
        f.puts "    option subnet-mask %s; " % netmask
        f.puts "    option broadcast-address %s;" % broadcast
        f.puts "    filename  \"pxelinux.0\";"
        f.puts "    next-server %s;" % tftp

        sites_for_vlan = if vlan < 10
                           [ site ]
                         else
                           root.sites(options)
                         end
        sites_for_vlan.each do |currentsite|
          currentsite.clusters(options).each do |cluster|
            cluster.nodes(options).each do |node|
              sitename=currentsite['uid']
              device = node['network_adapters'].find{|s| s['network_address'] =~ /^\w+-\d+\.\w+\.grid5000\.fr/}
              next if device.nil?
              hostname = device['network_address']
              next if hostname.nil?
              hostname_vlan = hostname.gsub(/^(\w+-\d+)(\..*)$/){$1+"-kavlan-"+vlan.to_s+$2}
              shortname_vlan = hostname_vlan.gsub(/^(\w+-\d+-\w+-\d+)(\..*)$/){$1}
              shortname = hostname.gsub(/^(\w+-\d+)(\..*)$/){$1}
              realsite = hostname.split(".")[1]
              begin
                vlan_ip = IPSocket::getaddress(hostname_vlan)
              rescue
                puts "WARN: Get address error: probably no DNS setup for vlan " +vlan.to_s+" on  site " + sitename + " , skip" if $opts.verbose;
                next
              end
              if device['mac'].nil? then
                if ref[shortname].nil? then
                  puts "WARN: mac undefined for host %s, skip" % hostname unless $opts.quiet
                  next
                else
                  mac = ref[shortname]['mac_eth0']
                end
              else
                mac = device['mac']
              end
              f.puts "   host %s {" % hostname_vlan
              f.puts "     hardware ethernet %s; " % mac
              f.puts "     option host-name \"%s\";" %  shortname_vlan
              f.puts "     option domain-name \"%s\.grid5000.fr\";" % realsite
              f.puts "     fixed-address %s;" % vlan_ip
              if vlan > 9
                # for global vlan, we need the local tftp server
                currenttftp = IPSocket::getaddress("kadeploy-server."+sitename+".grid5000.fr")
                f.puts "     next-server %s;" % currenttftp
              end
              f.puts "   }"
            end
          end
        end
        f.puts "}"
      end
    end
  end
end

(this script use restfully and ruby-ip gems)

Then, generate the configuration (replace VLANID and SITE by your current site and VLAN id), and copy it on the node:

frontend:

chmod +x ./gen_dhcpd_conf.rb

frontend:

gem install ruby-ip restfully --no-ri --no-rdoc --user-install

frontend:

./gen_dhcpd_conf.rb --site SITE --vlan-id VLANID

frontend:

scp dhcpd-kavlan-VLANID-SITE.conf root@`head -1 < nodefile`:

For user accounts, you need to specify your GEM_HOME directory because in the classical one, you won't be able to install the "ruby-ip". To make it possible, type :

frontend:

export GEM_HOME=~/.gem/ruby/2.3.0/

You have to disable the default DHCP server of the VLAN:

On the frontend

frontend:

kavlan -d

Now you have to install a DHCP server on the node (we assume the node is not yet in the job VLAN, or the vlan is routed and have access to internet for apt):

node:

apt-get install isc-dhcp-server

There may be an error after the installation : It's normal, you need to tell the DHCP server on which interface to listen to DHCP requests (replace "eno1" with the name of the interface on which the server should listen)

node:

sed -i s/INTERFACESv4=\"\"/INTERFACESv4=\"eno1\"/g /etc/default/isc-dhcp-server

You can now copy the generated configuration file and start the DHCP server :

node:

cp /root/dhcpd-kavlan-VLANID-SITE.conf /etc/dhcp/dhcpd.conf

node:

/etc/init.d/isc-dhcp-server restart

Then, in another shell, connect as root on a second node (or use kaconsole):

frontend:

ssh root@node-xx-kavlan-yy

And restart the network configuration:

node-dhcp-client:

systemctl restart networking

node-dhcp-client:

systemctl status networking

...
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr dhclient[2633]: DHCPREQUEST of 10.32.3.7 on eno1 to 255.255.255.255 port 67
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: DHCPDISCOVER on eno1 to 255.255.255.255 port 67 interval 10
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: DHCPREQUEST of 10.32.3.7 on eno1 to 255.255.255.255 port 67
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: DHCPOFFER of 10.32.3.7 from 10.32.3.6
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: DHCPACK of 10.32.3.7 from 10.32.3.6
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr dhclient[2633]: DHCPOFFER of 10.32.3.7 from 10.32.3.6
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr dhclient[2633]: DHCPACK of 10.32.3.7 from 10.32.3.6
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: bound to 10.32.3.7 -- renewal in 34620 seconds.
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr ifup[2618]: Sending network state change signal to nslcd...done.
Apr 12 17:05:34 suno-7-kavlan-4.sophia.grid5000.fr systemd[1]: Started Raise network interfaces.

on the dhcp server, check the logs:

node-dhcp-server:

tail /var/log/daemon.log

...
Apr 12 17:32:24 suno-6-kavlan-4 dhcpd: DHCPDISCOVER from 00:26:b9:3f:43:a1 via eno1
Apr 12 17:32:24 suno-6-kavlan-4 dhcpd: DHCPOFFER on 10.32.3.7 to 00:26:b9:3f:43:a1 via eno1
Apr 12 17:32:24 suno-6-kavlan-4 dhcpd: DHCPREQUEST for 10.32.3.7 (10.32.3.6) from 00:26:b9:3f:43:a1 via eno1
Apr 12 17:32:24 suno-6-kavlan-4 dhcpd: DHCPACK on 10.32.3.7 to 00:26:b9:3f:43:a1 via eno1

In the four last lines, you see that your own dhcp server has given an address to the other node.

DHCP and PXE

For your information, if you need to do a PXE boot, you must change the tftp server in the generated dhcpd configuration file:

node:

IP=`hostname -i`

node:

perl -i -pe "s/next-server .*/next-server $IP;/" /etc/dhcp/dhcpd.conf

(if there is no next-server configured, you must edit the file by hand and add a line like this:

next-server XX.XX.XX.XX ;

where XX.XX.XX.XX is the IP of your node (echo $IP).

Change the VLAN of your nodes manually

Put your nodes into the reserved VLAN

If you really want to change the VLAN manually, you can, but it's much simpler to change the vlan with kadeploy.

In order to change the VLAN of the nodes manually, you must reconfigure the network after the vlan has changed; but once the VLAN has changed, you can't connect to the node! An easy way to do this is to use the 'at' command (apt-get install at if it's not installed in your nodes)

We will use Taktuk to start remote commands on several nodes at once. In this example, we will use all the nodes. Since taktuk does not handle duplicate names in the nodefile, we must first remove duplicates.

First, we will use taktuk to install at on all nodes, then the taktuk command will simply launch the network reconfiguration in one minute. Finally, we set the VLAN of all our nodes.

As we will change the network configuration of nodes, we will use an isolated kavlan (a.k.a. kavlan-local) to not interfer with the rest of Grid'5000 network.

frontend:

oarsub -t deploy -l {"type='kavlan-local'"}/vlan=1,walltime=2 -I

frontend:

kavlan -V | tee myvlan

frontend:

oarsub -t deploy -l nodes=2 -I

frontend:

kadeploy3 -e debian10-x64-base -k -f $OAR_FILE_NODES

frontend:

taktuk -s -l root -f $OAR_FILE_NODES broadcast exec [ "apt-get update; apt-get --yes install at" ]

frontend:

taktuk -s -l root -f $OAR_FILE_NODES broadcast exec [ "echo '/etc/init.d/networking restart'| at now + 1 minute " ]

frontend:

kavlan -i `cat myvlan` -s -f $OAR_FILE_NODES

All nodes are configured in the vlan 2. In one minute, your nodes will renegotiate their IP addresses and will be available inside the VLAN, you can connect to each of them using kaconsole or ssh (as we use a kavlan-local, you must connect to the gateway of that kavlan first):

frontend:

ssh kavlan-`cat myvlan`

kavlan-VLANID:

ssh root@suno-30-kavlan-`cat myvlan`

You can use the ip neigh command to see the known hosts in your LAN; you should only see IPs in the 192.168.66.0/24 subnet

node:

ip neigh

192.168.223.254 dev eno1 lladdr 02:00:00:00:01:02 REACHABLE

You should be able to ping another of your host inside your VLAN

node:

ping -c 3 suno-42-kavlan-2

64 bytes from 192.168.211.42: icmp_req=1 ttl=64 time=0.141 ms
64 bytes from 192.168.211.42: icmp_req=2 ttl=64 time=0.166 ms
64 bytes from 192.168.211.42: icmp_req=3 ttl=64 time=0.165 ms

--- suno-42-kavlan-2.sophia.grid5000.fr ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 1998ms
rtt min/avg/max/mdev = 0.141/0.157/0.166/0.015 ms

Put your nodes back into the default VLAN

First, get put the list of your nodes name with vlan in a file:

frontend:

uniq $OAR_NODEFILE > mynodes

frontend:

sed "s/\([^.]*\)\(.*\)/\1-kavlan-`cat myvlan`\2/" mynodes > mynodes-vlan

Don't forget to first start the network restarting command, but this time, we need to run ssh from the kavlan gateway, and not the frontend :

frontend:

ssh kavlan-VLANID

kavlan-VLANID:

NODE in $(cat mynodes-vlan); do ssh root@$NODE "echo '/etc/init.d/networking restart' | at now + 1 minute "; done;

Then you can put your nodes back in the default VLAN:

frontend:

kavlan -s -i DEFAULT -f $OAR_NODEFILE

You should be able to ping your nodes:

frontend:

for i in `uniq $OAR_NODEFILE`; do ping -c 1 $i; done

Another way to put back nodes into the default VLAN is to change the vlan and then kareboot the nodes.

frontend:

kavlan -s -i DEFAULT -f $OAR_NODEFILE

frontend:

kareboot3 -f $OAR_NODEFILE -r simple

KaVLAN VPN

This feature allows users to build a Virtual Private Network (VPN) between a KaVLAN network and the outside world. Hence, it is possible to interconnect Grid'5000 nodes with any external network (from the user's laptop to the Internet), bypassing Grid'5000 network isolation.

	Warning
	This is an advanced feature. It requires a good understanding of KaVLAN, VPNs and networking in Linux

Some information:

• This service currently uses SSH VPN at Layer-2 (Ethernet level). Since that VPN is built on top of TCP protocol, you should not expect high network performance.
• The VPN requires two end points (or gateway) to be interconnected. On Grid'5000 side, VPN gateways are installed on kavlan-{1,2,3}.<site>.grid5000.fr servers.
• On the user's side (outside of Grid'5000), a GNU/Linux system with root privileges is required, to act as the user's gateway.
• On Grid'5000, the user must reserve a non-routed local kavlan network (the VPN only works with this kind of kavlan). Grid'5000 nodes must be switched into that kavlan to be accessible through the VPN.

The VPN is initiated from the user's gateway machine using a SSH connection to the appropriate kavlan-X server (which depends on kavlan network previously reserved). To enable VPN, SSH "-w" options must be used to connect to remote tap0 on kavlan-X server, with VPN tunnel configured in Ethernet mode. See ssh and ssh_config manpages for more information about those options.

Example, with KaVLAN network "1" at lyon :

• As root, create a virtual tap device that will be connected to your kavlan using SSH VPN. (Replace $USERNAME by your user name)

 laptop: sudo ip tuntap add dev tap0 mode tap user $USERNAME

• Assign an IP address to this interface.

 laptop: sudo ifconfig tap0 192.168.207.253/20

	Warning
	The IP address you choose must be inside the kavlan network, which depends on the kavlan number you are using. See Grid5000:Network#KaVLAN_networks

• Start the SSH VPN

 laptop: ssh -o Tunnel=ethernet -w 0:0 -N kavlan-1.lyon.g5k

If the command runs correctly, it should not output anything.

• Options description:
  • -o Tunnel=ethernet: Use an ethernet (layer 2) VPN
  • -w 0:0: Use interface tap0 on client side (first 0) and tap0 on server side (second 0, mandatory)
  • -N  : Do not execute a remote command
  • kavlan-1.lyon.g5k: Connect to lyon kavlan-1 gateway. Trailing .g5k assumes that you appropriately configured your ssh_config to connect to Grid5000 nodes using .g5k extension
• Client's tap0 interface is now connected to the kavlan network. You should be able to ping other nodes inside this network.

 laptop: ping 192.168.192.83
 PING 192.168.192.83 (192.168.192.83) 56(84) bytes of data.
 64 bytes from 192.168.192.83: icmp_req=1 ttl=64 time=82.7 ms
 64 bytes from 192.168.192.83: icmp_req=2 ttl=64 time=39.9 ms
 ...

	Warning
	DNS hostname resolution cannot be used here, as DNS servers are inside Grid'5000 network and this command is executed from your local workstation

Other usage

Using the API

Kavlan is also available through the API. Using the job and deploy API, you can, as with the command line tools, reverve nodes with vlan and deploy nodes into a vlan. If you want to manipulate VLAN directly through the API, you can do several things:

You can get the vlans you have reserved:

GET https://api.grid5000.fr/stable/sites/SITE/vlans/users/USER

You can get all the vlans available on a site:

GET https://api.grid5000.fr/stable/sites/SITE/vlans/

You can get the VLAN of nodes on a site:

GET https://api.grid5000.fr/stable/sites/SITE/vlans/nodes

You can print the VLAN of a list of nodes

POST https://api.grid5000.fr/stable/sites/SITE/vlans/nodes {\"nodes\": [ <list of node names>]}

You can change the VLAN of a list of nodes:

POST https://api.grid5000.fr/stable/sites/SITE/vlans/VLANID/ {\"nodes\": [ <list of node names>]}

You can start the dhcp server for the vlan

PUT https://api.grid5000.fr/stable/sites/SITE/vlans/VLANID/dhcpd {"action":"start"}

You can stop the dhcp server for the vlan

PUT https://api.grid5000.fr/stable/sites/SITE/vlans/VLANID/dhcpd {"action":"stop"}

Use a global VLAN

With a global VLAN, you can put nodes from several sites in the same VLAN

First reserve a global vlan on one site (here sophia) and 2 nodes on lille,sophia and lyon:

frontend:

oargridsub -t deploy -w 2:00:00 sophia:rdef="{\\\\\\\"type='kavlan-global'\\\\\\\"}/vlan=1+/nodes=2",lille:rdef=/nodes=2,lyon:rdef=/nodes=2 > oargrid.out

Get the oargrid Id and Job key from the output of oargridsub:

frontend:

export OAR_JOB_KEY_FILE=`grep "SSH KEY" oargrid.out | cut -f2 -d: | tr -d " "`

frontend:

export OARGRID_JOB_ID=`grep "Grid reservation id" oargrid.out | cut -f2 -d= | cut -d ' ' -f2`

Get the node list using oargridstat:

frontend:

oargridstat -w -l $OARGRID_JOB_ID | grep grid> ~/gridnodes

Then use kadeploy3 to deploy your image on all sites and change the VLAN:

frontend:

kadeploy3 -f gridnodes -a http://public.sophia.grid5000.fr/~nniclausse/openmx.dsc -k --multi-server -o ~/nodes.deployed --vlan 18

If you want to manipulate directly VLAN of a node, you have to run the kavlan command on the site where the node is, e.g. if you have reserved the global vlan located at sophia and want to put some nodes of lille into this vlan, you have to run kavlan -m nodename -i VLAN_GLOBAL_ID -s on lille site (or use the API with lille site in the URL).

How to use a local VLAN

In this section, we will describe the specificity of the local VLANs.

If you want to use local VLAN, you have to first connect on the gateway of the vlan. For this, once you have a running reservation on a local VLAN, you have a ssh accces to the gateway:

frontend:

ssh kavlan-<vlanid>

Then you can reach your nodes inside the VLAN. Another option is to use the kaconsole command.

(You can still use kadeploy to put your nodes in the VLAN in one step.)

Configure ssh to easily connect to nodes in a local VLAN

You can configure ssh to make the connection through the gateway transparent:

In order to transparently use ssh to acces to isolated nodes (local VLAN), you should add this to your .ssh/config file on the frontend:

Host *-*-kavlan-1 *-*-kavlan-1.*.grid5000.fr
    ProxyCommand ssh -a -x kavlan-1 nc -q 0 %h %p
Host *-*-kavlan-2 *-*-kavlan-2.*.grid5000.fr
    ProxyCommand ssh -a -x kavlan-2 nc -q 0 %h %p
Host *-*-kavlan-3 *-*-kavlan-3.*.grid5000.fr
    ProxyCommand ssh -a -x kavlan-3 nc -q 0 %h %p

Then you can simply use ssh <cluster>-<nodeid>-kavlan-<vlanid> to access the node , for example:

frontend:

ssh root@NODE-kavlan-VLANID

A simple multi NICs example

We show here how to reserve and configure multiple Ethernet network interfaces.

First we reserve a deploy job, with 2 nodes and 2 vlans:

frontend:

oarsub oarsub -I -t deploy -l {"eth_count > 1 and cluster = 'cluster_name'"}/nodes=2,{"type='kavlan'"}/vlan=2,walltime=02:00:00

Then we deploy the wanted environment:

frontend:

kadeploy3 -f $OAR_NODEFILE -k -e debian10-x64-nfs

See cluster section to know which Ethernet interfaces can be used. For exemple, on paranoia (Rennes), eth1/enp3s0f1 and eth2/eno1 are cabled. Use I=1 and J=2 for paranoia.

Get node name with interfaces:

frontend:

uniq $OAR_FILE_NODES | sed -e 's/\([^\.]*\)\(.*\)/\1-ethI\2/' > nodes_second_int

frontend:

uniq $OAR_FILE_NODES | sed -e 's/\([^\.]*\)\(.*\)/\1-ethJ\2/' > nodes_third_int

Show vlans number:

frontend:

kavlan -V

Put interfaces on the two different vlan:

frontend:

kavlan -i vlan1 -s -f nodes_second_int

frontend:

kavlan -i vlan2 -s -f nodes_third_int

Get ip on second and third interface :

frontend:

uniq $OAR_NODEFILE | taktuk -d -1 -l root -f - broadcast exec [ 'dhclient enp3s0f1' ]

frontend:

uniq $OAR_NODEFILE | taktuk -d -1 -l root -f - broadcast exec [ 'dhclient eno1' ]

At this moment your node should have 3 IP:

node:

ip a

root@paranoia-8:~# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1
    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: eno1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether f0:4d:a2:73:ce:3d brd ff:ff:ff:ff:ff:ff
    inet 10.24.70.8/18 brd 10.24.127.255 scope global eno1
       valid_lft forever preferred_lft forever
    inet6 fe80::f24d:a2ff:fe73:ce3d/64 scope link
       valid_lft forever preferred_lft forever
3: eno2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group default qlen 1000
    link/ether f0:4d:a2:73:ce:3e brd ff:ff:ff:ff:ff:ff
4: enp3s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether a0:36:9f:28:a9:18 brd ff:ff:ff:ff:ff:ff
    inet 172.16.100.8/20 brd 172.16.111.255 scope global enp3s0f0
       valid_lft forever preferred_lft forever
    inet6 fe80::a236:9fff:fe28:a918/64 scope link
       valid_lft forever preferred_lft forever
5: enp3s0f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group default qlen 1000
    link/ether a0:36:9f:28:a9:1a brd ff:ff:ff:ff:ff:ff
    inet 10.24.7.8/18 brd 10.24.63.255 scope global enp3s0f1
       valid_lft forever preferred_lft forever
    inet6 fe80::a236:9fff:fe28:a91a/64 scope link
       valid_lft forever preferred_lft forever