tyvm, David Boyes, et al:
Some progress:
I noticed Level 1 TCPIP did not have a NICDEF/SPECIAL for VSWITCH which
explains why Level 2 TCPIP cannot connect to Level 1 TCPIP via VSWITCH..
It had a SPECIAL for OSALAN, but OSALAN was not DEFINED in Level 1.
So I defined OSALAN in Level 1, coupled Level 2's 9112 , attached
9112-9114 in TCPIP Level 2's PROFILE EXEC, brought up TCPIP Level 2 and
finally got some connectivity:
DTCQDI001I QDIO device OSA9012 device number 9012:
DTCQDI007I Enabled for QDIO data transfers
DTCQDI001I QDIO device OSA9112 device number 9112:
DTCQDI007I Enabled for QDIO data transfers
DTCOSD238I ToOsd: IPv4 multicast support enabled for OSA9012
DTCOSD238I ToOsd: IPv4 multicast support enabled for OSA9112
DTCOSD319I ProcessSetArpCache: Supported for device OSA9012
DTCOSD319I ProcessSetArpCache: Supported for device OSA9112
DTCOSD341I Obtained MAC address 020000000008 for device OSA9012
DTCOSD341I Obtained MAC address 020000000009 for device OSA9112
DTCOSD246I OSD device OSA9012: Assigned IPv4 address 172.28.3.140
DTCOSD234I ToOsd: TCPIP host is not set as a router for port OSA9012 f
DTCOSD246I OSD device OSA9112: Assigned IPv4 address 172.28.3.240
DTCOSD234I ToOsd: TCPIP host is not set as a router for port OSA2 for
DTCOSD246I OSD device OSA9112: Assigned IPv4 address 10.13.13.26
But I am missing a DEFAULT Gateway (see below) and my subnet addresses and
masks are all wrong.
I suspect this is an MPROUTE problem.
I still cannot ping Level 1 from Level 2 and vv.
netstat gate
VM TCP/IP Netstat Level 540 TCP/IP Server Name: TCPIP
Path MTU Discovery Aging Interval: 10 Minutes
Known IPv4 gateways:
Subnet Address Subnet Mask FirstHop Flgs PktSz Metric Link
-------------- ----------- -------- ---- ----- ------ ------
172.28.0.0 255.255.0.0 <direct> U 576 1 OSA9112LNK
172.28.0.0 255.255.0.0 <direct> U 576 1 OSA9012LNK
Known IPv6 gateways: None
It should look something like my Level 1:
netstat gate
VM TCP/IP Netstat Level 540 TCP/IP Server Name: TCPIP
Path MTU Discovery Aging Interval: 10 Minutes
Known IPv4 gateways:
Subnet Address Subnet Mask FirstHop Flgs PktSz Metric Link
-------------- ----------- -------- ---- ----- ------ ------
Default <none> 172.28.3.129 UG 1492 7 OSA9000LNK
172.28.3.128 255.255.255.192 <direct> U 1492 1 OSA9000LNK
172.28.3.192 255.255.255.192 <direct> U 1492 1 OSA9100LNK
Known IPv6 gateways: None
MPROUTE does not find matches on ITF and RIP:
EZZ7883I Processing interface from stack, address 10.13.13.26, name
VIRLNK1, ind
ex 1, flags 4041
EZZ7871I No matching interface statements for 10.13.13.26 (VIRLNK1)
EZZ7883I Processing interface from stack, address 172.28.3.140, name
OSA9012LNK,
index 2, flags 463
EZZ7871I No matching interface statements for 172.28.3.140 (OSA9012LNK)
EZZ7883I Processing interface from stack, address 172.28.3.240, name
OSA9112LNK,
index 3, flags 463
EZZ7871I No matching interface statements for 172.28.3.240 (OSA9112LNK)
EZZ8023I The RIP routing protocol is Enabled
I added the subnet mask to my HOME statements but I get:
17:25:37 DTCPDO151E Subnet mask supplied by MPRoute (255.255.0.0) for link
OSA91
12LNK does not match subnet mask specified on the HO
ME statement (255.255.255.192).
INTERFACE NAME=VIRLNK1
IP_ADDRESS=10.13.13.26
SUBNET_MASK=255.255.255.252
MTU=1492;
RIP_INTERFACE NAME=OSA9012LNK
IP_ADDRESS=172.28.3.140
SUBNET_MASK=255.255.255.192
RIPV2=YES
RECEIVE_RIP=YES
RECEIVE_DYNAMIC_NETS=YES
RECEIVE_DYNAMIC_SUBNETS=YES
RECEIVE_DYNAMIC_HOSTS=YES
MTU=1492;
RIP_INTERFACE NAME=OSA9112LNK
IP_ADDRESS=172.28.3.240
SUBNET_MASK=255.255.255.192
RIPV2=YES
RECEIVE_RIP=YES
RECEIVE_DYNAMIC_NETS=YES
Any ideas?
----- Forwarded by George Henke/NYLIC on 10/13/2010 04:17 PM -----
David Boyes <[email protected]>
Sent by: The IBM z/VM Operating System <[email protected]>
10/13/2010 02:45 PM
Please respond to
The IBM z/VM Operating System <[email protected]>
To
[email protected]
cc
Subject
Re: Cannot TELNET to Level 2
Let¹s back up a bit.
What you are essentially doing is creating a setup like this:
Company network --> vswitch -> level 1 network stack
+-----> level 2 network stack
The TCP stacks (both level 1 and level 2) need:
1) unique IP addressess
2) the same network mask
3) the same default gateway
4) NICs attached to the same VSWITCH in the same mode (both ETHERNET or
both
IP)
The IP addresses should be from the same network range if they are
attached
in this form.
What it looks like is that you don't have a default route yet, or
something
went wrong when you coupled the level 2 stack to the vswitch. Since you
can't reach the level 1 host, I'd bet on the latter.
On 10/13/10 1:09 PM, "George Henke/NYLIC" <[email protected]>
wrote:
>
> I have NICDEFed to my level 1 switch from my level 2 machine (ZVM54SVM)
and
> can display my level 2 switch now with "q lan" from Level 2:
>
> Adapter Owner: VLINUX3 NIC: D000.P00 Name: LNXVSW1
> RX Packets: 341673 Discarded: 0 Errors: 0
> TX Packets: 24165 Discarded: 0 Errors: 0
> RX Bytes: 23039266 TX Bytes: 7154034
> Device: D002 Unit: 002 Role: DATA vPort: 0067 Index:
0067
> Options: Broadcast Multicast IPv6 IPv4 VLAN
> Unicast IP Addresses:
> 172.28.3.137 MAC: 02-00-00-00-00-06
> FE80::200:0:100:6 MAC: 02-00-00-00-00-06 Local
> Multicast IP Addresses:
> 224.0.0.1 MAC: 01-00-5E-00-00-01
> 224.0.0.251 MAC: 01-00-5E-00-00-FB
> FF02::1 MAC: 33-33-00-00-00-01
> FF02::FB MAC: 33-33-00-00-00-FB
> Adapter Owner: ZVM54SVM NIC: 9012.P00 Name: UNASSIGNED
This doesn't look right. Did you use a port name when you did the define
for
the VLINUX3 adapter?
> Also I do not see much detail for my Level 2 machine in the "q lan"
above.
Yeah. That's definitely uncool.
> Can't I just bypass VSWITCH and OSALAN and just attach the same real
device
> OSA device address from Level 1 to Level 2 to a unique IP address?
No. See the "drawing" above. The two systems can't have the same IP
addresses because they ARE on the same LAN.
Suggestion:
On the level 2 machine, change SYSTEM NETID on MAINT 190 to use a fake CPU
serial number for the 2nd level system. Call it LVL2TEST in SYSTEM NETID.
On 1st level, log into the userid for your 2nd level system and use SET
CPUID to set the virtual CPUID for that machine to the "fake" one.
In the level 2 TCPIP configuration: duplicate the link, home, etc lines
and
put LVL2TEST: in front of one copy of the lines.
Change the lines in the level 2 TCPIP configuration to give the level 2
guest a new IP address in the same network as the level 1 system (eg if
level 1 is 172.16.82.x, the level 2 system should be 172.16.82.y).
The default route and other configuraiton information should be identical
to
level 1 (ie, if you're using MPROUTE, the configuraiton should be the
same).
Re-IPL level 2 system. You should see the system identifier in the bottom
right corner as LVL2TEST. When the TCPIP machine starts up, it should
select the new IP address.
That way you can connect directly to either system during testing, but
when
you go live with level 2, you'll get the normal ip address of the level 1
system without changing anything.
