HP XC System Software Installation Guide Version 2.1

_________________________ Note _________________________
If the ---oldmp option was used on the discover command line, it is assumed
that all management processors (MPs) h ave their IP addresses set statically, and
therefore are not subject to this step in the discovery process.
If so me console p orts are not configured to use DHCP, they will not be discovered. Therefore,
the first item to verify is whether or not the non-discovered console ports are configu red to
use DHCP.
If all console ports are properly configured to use DHCP, verify the wiring against the
descriptions in the HP XC Hardware Preparation Guide. It is important to point out that because
of h a rdware lim it ations in the ProCurve switch, empty ports are included in the port layout.
E.2.4 Some Console Por ts Have Not Obtained Their IP Address
Use the following procedure to determine why all console ports have been discovered, but some
have not obtained their IP address after a reasonable time:
1. View the system log file:
# tail -f /var/log/messages
2. Look for instances where a netwo
rk component issues a DHCPREQUEST, the head node
sends back a DHCPOFFER, but the c
onsole port does not send back a correspondin g
DHCPACK. For example:
Sep 27 09:47:46 src_s@n4 dhcpd: DHCPDISCOVER from 00:30:6e:22:66:02 via eth1
Sep 27 09:47:46 src_s@n4 dhcpd: DHCPOFFER on 172.21.0.3 to 00:30:6e:22:66:02 via eth1
eth1
3. If this is the case, cycle power on the affected nodes, or if more than a few nodes are
affected, cycle power on the nodes in the cabinet.
It is important to note that just turning the pow er off o n the node is not sufficient because
power is present on the console port whenever AC power is present. To totally reset a
node, you must remove power by unplugging the node or by using the circuit breakers on
the cabin et power distribution unit (PDU).
E.2.5 Not All Nodes Are Discovered
In much the same way as the c onsole ports are queried, the discover process queries the
ProCurve switches to obtain the MAC address of a node. To do th is, the nodes must be
configured to boot off the network interface that is connected to the Root Administr ation
Switch. If the node is no t configured to network boot, it is likely that the node will not b e
properly discovered.
The discover process u ses the num ber o f nod es as a checkpo int t o determ in e whether o r not it
has discovered all the nodes i n the system. As a r esult , the system will check the po rts on a
switch for an open port. If one is found, this does not necessarily indicate a problem but rather
where the nodes on the particular switch end.
When this non-responding port is found, the discover process proceeds to the next switc
hinthe
list. It will do this until it discovers all nodes or runs out of ports. As a result, it is im
portant that
all nodes in the cluster be available during the discovery process.
To determine w h a t node failed in the discover process, examine the output of the discover
command when i t is parsing the switch outp ut t o gather the n ode input. For example, assume
that the f ollow ing was ou tpu t during the discovery process:
.
.
.
Switch 172.20.65.3 port 4 ... Node Found
Switch 172.20.65.3 port 5 ... Node Found
Discovery Flowchart and Tr oubleshooting E-3