:::HMC DLPAR ISSUES SORTED OUT:::

1) From *VIO can it ping vhmc or hmc ?

ping -S <VIO IP> <HMC IP>

.

2) From vhmc or hmc can it ping *VIO ?

ping -I <HMC IP>  <VIO IP>

.

Check if telnet works via port 657:

3) From *VIO:

telnet <HMC IP> 657

netstat -Aan | grep 657

.

4) From vhmc or hmc:

telnet <VIO IP> 

netstat -anp | grep 657

.

5) If firewall exists between *VI0 and vhmc or hmc, please check that port 657 TCP/UDP is opened bi-directional on the firewall.

.

Here we need pesh pw's for running commands on hmc. You need to be root. (you have to ask IBM vendor to provide Pesh password)

chhmcusr -u hscpe -t passwd

ssh hscpe@localhost

then

pesh serial-number-of-HMC-machine

As you can see from the below the telnet from HMC to *VIO is showing the port 657 TCP is blocked.

.

[root@vhmc ~] # telnet <VIO IP> 657

Connection closed by foreign host

.

That means port 657 TCP/UDP is blocked on firewall/switch/router between vhm and *VIO.

.

So your next step is to check with your firewall team and get the port 657 TCP/UDP opened bi-directional.

IF we  found that both *VIO1 and *VIO2 have duplicate nodeid's

This can cause unpredictable results and can cause a good RMC connection to bounce between lpars. 

VIO1:

#cat ct_node_id

1709efca1ed5b6c8

VIO2:

#cat ct_node_id

1709efca1ed5b6c8

So to correct this we need to generate a new nodeid on *VIO1.

Please do the following on *VIO1:

1) /usr/sbin/rsct/install/bin/recfgct

2) /usr/sbin/rsct/bin/rmcctrl -p

/usr/sbin/rsct/bin/rmcctrl -z

/usr/sbin/rsct/bin/rmcctrl -A

/usr/sbin/rsct/bin/rmcctrl -p

lssrc -a | grep rsct

lsrsrc IBM.MCP

[]

# netstat -an | grep 657

# lspartition -dlpar | grep -A 1 <lpar_ipaddress>

Then DLPAR should be working fine now..