I have Linux host with libvirt/kvm virtualization, VMs needs the "real" static IP addresses, so I decided to setup bridged network. I make br0 on the host and in the VMs properties I set source device: Host device vnet0(Bridge 'br0').
For example, my br0 have ip 192.168.1.1 and one of the VM have 192.168.1.5
Everything works pretty well, but then I connect to the virtual machine, the client address detects as 192.168.1.1. Also, all the HTTP requests comes from 192.168.1.1.
Q: Is it my mistake, some sort of misconfiguration? How can VM get the real IPs of the clients?
Let me try to answer based on what i infer from your question:
Since you want to assign routable IP addresses to the VMs,
Option 1: Add the host physical ethernet interface to the vswitch (aka vswitch uplink). Further, for all the VM ethernet interface, assign IP address in the same subnet in which the physical ethernet interface's IP belongs. Alternatively, if DHCP server is running in the same broadcast domain (subnet), the VMs would get the IP from the DHCP server if the interfaces are configured to get IP via DHCP
Option 2: Create the vswitch and assign X.Y.Z.1 IP to the vswitch (br0). Also enable IP forwarding in the host. Now you can assign IPs from the same subnet to the VM ethernet interfaces. Alternatively, you can run DHCP (e.g. DNSmasq) on br0 and assign IPs to the VM interfaces
Is it my mistake, some sort of misconfiguration? How can VM get the real IPs of the clients?
If you are connecting from the host on which your vms are running, then they are getting the real IP address. Your host shares a network (192.168.1.0/24 or similar, apparently) with the virtual machines. When you connect to your virtual machines from your host, the source address if 192.168.1.1.
If you are connecting from elsewhere on your network, you would need to provide us with more details of your configuration.
Related
Some background:
I'm attempting to set up a pentesting network with a handful of virtual machines for the SANS 560 (Network Penetration Testing and Ethical Hacking) course, but I'm having an issue with the network configuration.
To paint a picture of the network (at least how it's intended to be):
My home router (connected to the internet, also the gateway for all other machines on the network) IP is 192.168.0.1/24, with all other machines on the network in the 255.255.255.0 subnet
As per the course notes, I should be setting up all my virtual machines with bridged adapters on the 10.10.X.X/16 subnet - with Linux machines on 10.10.75.X/16, Windows guest machines on 10.10.76.X/16, and my "host" (also a VM running Windows) machine on 10.10.78.1/16
My question:
How (assuming it's possible) do I configure my host machine (with new new IP 10.10.78.1/16) to be able to talk to the other guest machines (Virtual machines) while also being able to connect to the internet?
I've tried setting up a static route to use the new IP as the gateway (seeing as the router is on a different subnet):
route ADD 192.168.0.0 MASK 255.255.255.0 10.10.78.1 (192.168.0.0 is the destination, obviously the mask is 255.255.255.0, and the gateway is 10.10.78.1) - it didn't work (all I get is Destination Host Unreachable)
Do I need to have two interfaces on this Windows machine (i.e. one configured as 10.10.78.1/16 to talk to the other VMs, and another configured as 192.168.0.X/24 to access the internet) to make this configuration possible?
I understand it's not how a network would be set up typically, so please let me know if you need me to clarify or provide more information.
I found a solution that seems to work.
Again, for context, below is a list of the machines on the network:
Name | Adapter type | IP | Static routes?
============================================================================
Windows VM1 | Bridged | 192.168.0.11/24 | Nil
| Bridged | 10.10.78.1/16 | Yes, see below
-------------|----------------|---------------------|-----------------------
Windows VM2 | Bridged | 10.10.76.1/16 | Yes, see below
-------------|----------------|---------------------|-----------------------
Linux VM3 | Bridged | 10.10.75.1/16 | Nil
Static routes:
Static routes for VM1:
Note: In the adapter settings for 192.168.0.11/24, I set the default gateway as the IP for my internet router (192.168.0.1), and the netmask as 255.255.255.0
Note: In the adapter settings for 10.10.78.1/16, I left the default gateway blank (as it gets set when adding the static route), and the netmask as 255.255.0.0
route -P ADD 10.10.0.0 MASK 255.255.0.0 192.168.0.11 (must use -P so that the route persists between reboots)
Static routes for VM2:
route -P ADD 10.10.0.0 MASK 255.255.0.0 10.10.76.1 (must use -P so that the route persists between reboots)
Note: you must run netsh advfirewall set allprofiles state off to allow the other VMs (including other Windows machines) on the 10.10.0.0/16 subnet to talk to this machine.
This configuration allows the following behaviour:
VM1 can initiate a connection with VM2 and VM3
Neither VM2 nor VM3 can initiate a connection with VM1
VM2 and VM3 can inter-communicate (i.e. can initiate connections with each other, in either direction)
Furthermore, this configuration should allow all of the VMs to communicate to the VPN that is setup for labs later on in the course, since they all have an adapter configured on the 10.10.x.x/16 network.
There are two solutions:
Add 10.10.0.0/16 to your router as a secondary IP subnet - if possible - or change the 192.168.0.0/24 range to 10.10.0.0/16.
Use another router to create the 10.10.0.0/16 subnet and connect it to 192.168.0.0/24 through on of its interfaces. On your Internet router, add a static route to 10.10.0.0/16. The router can be anything, a hardware router, a layer-3 switch, or a Windows/Linux machine with routing enabled.
A third approach - running both subnets in the same layer-2 segment connected by a router-on-a-stick - doesn't really cut it for the purpose.
Edit: The route in your question is the wrong way - assuming your inter-subnet router uses 192.168.0.99 and 10.10.78.1, on your Internet router, add route 10.10.0.0/16 -> 192.168.0.99 and on the new subnet use 10.10.78.1 as the default gateway.
How to access host files from VM Oracle enterprise linux 6.3 (Virtual box).
It is not connected to internet and it has only terminal.
I downloaded Winscp, but it needs hostname(i.e IPaddress of VM), as it is not connected to internet, IP is not available. Is there any other alternative solution to access host files.
Please let me know which networking type you used for RHEL 6.3.
If you used NAT, then add port forwarding rule (host port 22 and guest port 22). and use IP address 127.0.0.1
If you used host-only adapter the it will have private network IP address. Use that private IP address.
If you used bridged then it will have a public IP address.
Please refer my blog for further details:
ref: http://trainenv.blogspot.in/2016/03/virtual-networking.html
I'm facing some problems trying to directly connect docker containers to the network of the host.
The configuration is as follows
One host has one interface (eth0) in the subnet, say, 10.0.15.0/24. The IP on eth0 is 10.0.15.5/24.
I customized the docker0 bridge to use a subnet within the subnet available from eth0, namely 10.0.15.64/26. So docker can use IPs from this /26 to give to containers, and I want the containers to be directly accessible from the rest of the network. The docker bridge also has an IP set, namely 10.0.15.65/26.
When a containers is created, it gets an IP, say 10.0.15.66/26. Now, I did some test with pinging:
anything on the network can ping 10.0.15.5 (eth0 of host)
anything on the network can ping 10.0.15.65 (docker0 bridge of host)
host can ping 10.0.15.66 (ip of container)
container can ping anything on the network
anything other than the host can not ping the container at 10.0.15.66
IP forwarding is turned on
[root#HOSTNAME~]# cat /proc/sys/net/ipv4/ip_forward
1
What am I missing here?
The containers connected to the docker0 bridge should be reachable from the network I think.
Expected behaviour
Containers should be pingable from anywhere on the network, just like the docker0 bridge etc.
Any thoughts or help would be greatly appreciated!
Finally figured out why it wasn't working for us.
The machine I was running the docker container in, was a VM on a hypervisor. The hypervisor only accepts one MAC address from the NIC attached to the VM. In other words, the NIC in the VM was not set to promiscuous mode.
What I did to work around this issue was just use a bare metal machine. Another solution would be to manually set the NIC to promiscuous mode, so it accepts all packets, instead of just the packets for it's own MAC.
i can see that virtual ip for a service can be fixed, but it has nothing about the outbound ip and port . in other words -- When i send something like a response from my virtual machines do they have the same source ip and port or have different ips based to the virtual machines.
Your VMs will be presented to the internet using the VIP you reserve. The VIP is on the wrapping Cloud Service which also provides load balancing for the inbound ports. See here for more information: http://blog.kloud.com.au/2014/07/18/static-dip-request-vip-reservation-on-microsoft-azure/
Host Machine is Windows 7
and I am using Oracle Virtual Box ( GUEST OS - FEDORA 20). I have enabled bridged connection in between host machine and guest machine. (Did it by selecting bridge connction inside virtual box manager)
Now I am a bit confused about IP addresses.
On my FEDORA guest machine when I do ifconfig. It shows me 2 interfaces. lo (localloop) and p2p1 with an IP address 192.168.0.12.
When I use firefox inside fedora is shows me some ip address which is same ip address as my host machine. (I checked using whats my ip address dot com)
Now I have a few questions
a) How come my guest OS and Host OS have same IP address ?
b) From where does interface p2p1 is getting an address 192.168.0.12 ? Is it assigned by my Wifi router ?
If yes, then is it possible to do ssh to 192.168.0.12 to reach my virtual machine from other machines connected to same router ? (provided if i enable ssh on my virtual machine)
c) Since it's an bridge network, How can I ssh or telnet my host windows machine from virtual machine ?
When I use firefox inside fedora is shows me some ip address which is same ip address as my host machine.
No. It shows the public IP address of your DSL router which is the same for your host and the virtual machine, as any traffic to the internet is going over the router. You need to understand that first.
When you have enabled "bridged connection" the vm is in the same private subnet as your host machine. The IP address 192.168.0.12 has been assigned by the local DHCP server, which is running on your router.
Since both, the vm and your host computer are in the same network you can ssh into your host computer from the vm using your host computer's IP address, which will be 192.168.0.???.