Difference between revisions of "MDs Do Not Network (PXE) Boot"

From LinuxMCE
Jump to: navigation, search
(Media Directors to not network boot - PXE)
 
Line 3: Line 3:
 
There are many reasons why this may happen:
 
There are many reasons why this may happen:
  
- Using a manageable switch with STP, where the ports are Blocking (eg Cisco enterprise switches)
+
=Using a manageable switch with STP, where the ports are Blocking (eg Cisco enterprise switches)=
- Media Director is not on the internal (LMCE) LAN
+
- DHCP server is not operational
+
- DHCP server has run out of leases
+
- Using VLANs
+
- Second DHCP server is present on the same LAN segment
+
- TFTP server is not operational
+
- Core/Hybrid is not on, operational (crashed) or unreachable on the network
+
 
+
 
+
Using a manageable switch with STP
+
  
 
Many switches use Spanning Tree Protocol (STP) which by default sets a port to "Blocking" and other non-Forwarding states for the first 50 seconds after a computer, connected to it, comes online. This is to allow time for the switch to confirm that there is no loop in your network related to that port, and during this period it will not pass any packets onto the rest of the LAN. This means your MD's network (PXE) boot process will almost certainly timeout and fail before the port goes into its Forwarding state. Power off/on, and in most cases, rebooting will simply restart the STP timer and end in the same timeout.
 
Many switches use Spanning Tree Protocol (STP) which by default sets a port to "Blocking" and other non-Forwarding states for the first 50 seconds after a computer, connected to it, comes online. This is to allow time for the switch to confirm that there is no loop in your network related to that port, and during this period it will not pass any packets onto the rest of the LAN. This means your MD's network (PXE) boot process will almost certainly timeout and fail before the port goes into its Forwarding state. Power off/on, and in most cases, rebooting will simply restart the STP timer and end in the same timeout.
Line 28: Line 18:
 
Then disconnect your telnet session.
 
Then disconnect your telnet session.
  
Media Director is not on the internal (LMCE) LAN
+
=Media Director is not on the internal (LMCE) LAN=
DHCP server is not operational
+
=DHCP server is not operational=
DHCP server has run out of leases
+
=DHCP server has run out of leases=
Using VLANs
+
=Using VLANs=
Second DHCP server is present on the same LAN segment
+
=Second DHCP server is present on the same LAN segment=
TFTP server is not operational
+
=TFTP server is not operational=
Core/Hybrid is not on, operational (crashed) or unreachable on the network
+
=Core/Hybrid is not on, operational (crashed) or unreachable on the network=

Revision as of 23:46, 4 January 2008

You have a working Core/Hybrid and network - you start a new PC on the LMCE (internal) network, and you tell it to network boot (PXE) so that you can use it was an Media Director. The boot process starts but the PC does not successfully boot as a Media Director.

There are many reasons why this may happen:

Using a manageable switch with STP, where the ports are Blocking (eg Cisco enterprise switches)

Many switches use Spanning Tree Protocol (STP) which by default sets a port to "Blocking" and other non-Forwarding states for the first 50 seconds after a computer, connected to it, comes online. This is to allow time for the switch to confirm that there is no loop in your network related to that port, and during this period it will not pass any packets onto the rest of the LAN. This means your MD's network (PXE) boot process will almost certainly timeout and fail before the port goes into its Forwarding state. Power off/on, and in most cases, rebooting will simply restart the STP timer and end in the same timeout.

Assuming that you definitely do not have any loops in your network, you can bypass this checking process by configuring that port. Never do this to a port that connects to another switch - setting up a loop between switches or to the same switch, without this check can paralyse your network. For a Cisco switch, connect to your switch's management IP address with telnet and type:

enable config t interface e1/0 <- change e1/0 to the port's blade and port number portfast enable exit write

Then disconnect your telnet session.

Media Director is not on the internal (LMCE) LAN

DHCP server is not operational

DHCP server has run out of leases

Using VLANs

Second DHCP server is present on the same LAN segment

TFTP server is not operational

Core/Hybrid is not on, operational (crashed) or unreachable on the network