** Description changed:

  Context is the QA lab
  
  In one VM I have the region-controller (RC) where:
  
  10.98.0.90 is the api/webui
  192.168.21.5 is the RC's provisioning server
- 
  
  On another VM I have a cluster-controller (CC) where
  
  192.168.20.5 is the CC's provisioning server
  10.98.0.91 is the interface connected to the RC network
  
  The 21.x and 20.x networks don't have access to each other.
  
  When I install the CC, I set up the API to http://10.98.0.90/MAAS
  
  When the CC is registered to the RC and and dhcpd.conf file is written,
  I get this:
  
  subnet 192.168.20.0 netmask 255.255.255.0 {
-        next-server 192.168.21.5;
-        filename "pxelinux.0";
-        option subnet-mask 255.255.255.0;
-        option broadcast-address 192.168.20.255;
-        option domain-name-servers 192.168.21.5;
-        option routers 192.168.20.1;
-        range dynamic-bootp 192.168.20.10 192.168.20.20;
+        next-server 192.168.21.5;
+        filename "pxelinux.0";
+        option subnet-mask 255.255.255.0;
+        option broadcast-address 192.168.20.255;
+        option domain-name-servers 192.168.21.5;
+        option routers 192.168.20.1;
+        range dynamic-bootp 192.168.20.10 192.168.20.20;
  }
  
  When the nodes in the 20.x network are booted, they can reach the DHCP server 
but can't reach the provisioning server.
  If I change the next-server value to 192.168.20.5, the nodes can't find the 
images.
  
  Looks like when the CC is registered, the images in the RC should be
  sent to the CC, so it can serve them to the nodes in its network.
+ 
+ 
+ = python-tx-tftp =
+ 
+ [Impact]
+ Enables the backend to know from what address the request came in on. This is 
an important fix related to MAAS, as it requires from what IP address the 
request came from in order to provide tftp correctly.
+ 
+ [Test Case]
+ 1. Install MAAS region in 1 machine.(Make sure iscsi is not running)
+ 2. Install MAAS cluster in a second machine. (ISCSI will be running managed 
by MAAS)
+ 3. Boot a client machine.
+ If the enlistment failed it is because the machine tried to access the region 
controller instead of the cluster controller. If the machine enlists successful 
we can be confident that the bug is resolved.
+ 
+ [Regression Potential]
+ Minimal. This simply passes a new argument and does not affect the code. This 
has been tested wwith MAAS and the MAAS team is committed to fix any issues for 
this package from this point forward.

-- 
You received this bug notification because you are a member of Ubuntu
Server Team, which is subscribed to maas in Ubuntu.
https://bugs.launchpad.net/bugs/1068843

Title:
  maas-cluster-controller doesn't have images for provisioning

To manage notifications about this bug go to:
https://bugs.launchpad.net/maas/+bug/1068843/+subscriptions

-- 
Ubuntu-server-bugs mailing list
Ubuntu-server-bugs@lists.ubuntu.com
Modify settings or unsubscribe at: 
https://lists.ubuntu.com/mailman/listinfo/ubuntu-server-bugs

Reply via email to