U6ID Principal

50 views
Skip to first unread message

Daivik Dave

unread,
May 4, 2018, 8:58:43 AM5/4/18
to Linux XIA
Hello , 
        I have started working on U6ID principal . Can i use 0x1b for U6ID principal type , If it is not allocated to any other principal ?

Regards
Daivik

Michel Machado

unread,
May 4, 2018, 2:30:42 PM5/4/18
to Daivik Dave, Linux XIA
Hi Daivik,

I just allocated 0x1b for U6ID for you. Thank you for requesting it. Just for reference, we keep all constants used in XIA on the following page: https://github.com/AltraMayor/XIA-for-Linux/wiki/Constants-used-in-XIA . 

--
You received this message because you are subscribed to the Google Groups "Linux XIA" group.
To unsubscribe from this group and stop receiving emails from it, send an email to linux-xia+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Daivik Dave

unread,
May 17, 2018, 10:46:34 AM5/17/18
to Linux XIA
Hello,
             I have a doubt , is struct fib_xid_u4id_local have one instance on a machine or it has instances corresponding to each entry in the local table , also there is a socket in the context and also in the local u4id struct , do these sockets have different purposes (i have some confusion here.)
Thanks 
Daivik

Michel Machado

unread,
May 18, 2018, 4:06:44 PM5/18/18
to Daivik Dave, Linux XIA
There is an instance of struct fib_xid_u4id_local for each entry in the local table.

Each local entry has a listening socket, so the TCP/IP stack receives the packet and passes it to the XIA stack.

The socket is the context is the socket XIA uses to transmit packets over TCP/IP.

--

Daivik Dave

unread,
May 22, 2018, 1:17:23 PM5/22/18
to Linux XIA
Hello,
               I am getting a no space error "Can't obtain a virtual XID type."  , Is this because  of the constant XIP_MAX_XID_TYPES set to 8 , exactly 8 xia principals are loaded in my OS . If this is the case , do i need to compile the kernel again ,if i increase it to 9?
    Thanks,
     Daivik

 

Michel Machado

unread,
May 22, 2018, 4:11:23 PM5/22/18
to Daivik Dave, Linux XIA
Try another free id for U6ID instead of 0x1b until it works. The pull request #16 is meant to solve this problem, but it's in our backlog. We'll try to clear this pull request before you are ready to merge your work.

--

Daivik Dave

unread,
May 25, 2018, 6:15:34 AM5/25/18
to Linux XIA
Hello,
         Is there any reference material i can use for IPv6 version of handle_skb_to ipv4 function?
Thanks
Daivik

Michel Machado

unread,
May 25, 2018, 11:48:42 AM5/25/18
to Daivik Dave, Linux XIA
I know it doesn't sound motivating, but the only reference material I'm aware of is the code of the kernel. Not only the XIA part, but the kernel itself is full of potential examples.

--

Daivik Dave

unread,
Jul 12, 2018, 1:59:46 AM7/12/18
to Linux XIA
Hello ,
         I have been trying to replicate the u4id experiment for u6id , and there are some problems .
1) The default IPv6 address assigned to eth0 interface is a Link local address and it cannot be used to bind to a socket .    
    2) The ip address assigned through ip add addr command gets successfully added to the interface and also gets added to a new route in the principal , but is not recognised by other containers . 
   3) I tried some commands specified in https://www.linux.com/learn/another-ipv6-crash-course-linux-real-ipv6-addresses-routing-name-services , an image attached to demonstrate the commands, But i am unable to ping between containers.
  
What should i do to assign Ipv6 address to the host and the containers in the star topology , so that the Ipv6 address are global and the containers are able to ping each other ?
Thanks 
Daivik
commands.jpg

Michel Machado

unread,
Jul 13, 2018, 9:58:07 AM7/13/18
to Daivik Dave, Linux XIA
1. Manually assign an IPv6 address and use it in the experiment.

2. It reads as you haven't made sure that the routing tables of the containers have been updated to reflect the new address. If you have updated the routing tables of the containers, look for small mistakes in them. We had a small mistake in the routing tables of the exemplifying demo (https://github.com/AltraMayor/XIA-for-Linux/wiki/An-exemplifying-demo) when we were building that experiment that took us quite some time to figure out.

3. The fact that you can't ping between the containers increases the chances of what I said in item 2 to be the root cause.

You need to set pairs of IPv6 addresses that are set up as local addresses between the central host and an outter point of the star. Like the IPv4 experiment does: 10.0.1.1 and 10.0.1.2, 10.0.2.1 and 10.0.2.2.

--

Daivik Dave

unread,
Jul 23, 2018, 8:04:09 AM7/23/18
to Linux XIA
Hello ,
        By assigning manual IPv6 addresses and routes on the host , The U6ID principal is now working . This is the link( https://drive.google.com/file/d/1KPirU-IrkbsZXxdkJyl71SA9nkpvmVgs/view?usp=sharing ) of the video demonstrating the U6ID principal . Currently Saurav Kumar is reviewing my code , and will submit pull request when the review and changes are complete .

Thanks
Daivik

Michel Machado

unread,
Jul 24, 2018, 11:04:33 AM7/24/18
to Daivik Dave, Linux XIA
Congrats on getting the U6ID principal working, Daivik.

--
Reply all
Reply to author
Forward
0 new messages