VMS cluster setup problem

[#1]

Hi,

I am trying to setup a two node VMS LAN cluster. One system is Alpha 4100 and other is DS-20 server. Both are running OVMS 8.3

After configuring the system for lan cluster, I restarted the system. Both the system comes up but the show cluster shows following output on both the node:

| SYSTEMS | MEMBERS |
|————————-——————————————+—————————|
| NODE | SOFTWARE | STATUS |
|————————+——————————————+—————————|
| WM6 | VMS V8.3 | MEMBER |
| WM3 | VMS V8.3 | BRK_NEW |
+————————-——————————————-—————————+


Could you help me out to solve this problem.

Below is the output of ifconfig -a from both the systems.

On WM6
$ ifconfig -a
IE0: flags=c43
*inet 16.138.73.61 netmask ffffff00 broadcast 16.138.73.255 ipmtu 1500

LO0: flags=100c89
inet 127.0.0.1 netmask ff000000 ipmtu 4096

TN0: flags=80

TN1: flags=80


On WM3
$ @SYS$COMMON:[SYSMGR]TCPIP$DEFINE_COMMANDS.COM;1
$ ifconfig -a
LO0: flags=100c89
inet 127.0.0.1 netmask ff000000 ipmtu 4096

TN0: flags=80

TN1: flags=80

WE0: flags=c43
*inet 16.138.73.58 netmask ffffff00 broadcast 16.138.73.255 ipmtu 1500
inet 16.138.73.57 netmask ffffff00 broadcast 16.138.73.255 ipmtu 1500

Please let me know if any further information is required to help.

Regards,
ajaydec

[Bart Zorn_1]

Hello ajaydec,

Yes, we need a lot more information. Actually, almost all information that you supplied is mostly irrelevant, because OpenVMS clustering does not (yet) use TCP/IP at all.

Please describe what you did so far to create the cluster. Then we may be able to help you.

The console output of the boot attempt of WM3 may be helpful as well.

Regards,

Bart

[Volker Halle]

ajaydec,

if both systems come up, but they don't see each other in SHOW CLUSTER, you've certainly done something wrong.

In a configuration like this, you would need to corretly set up VOTES and EXPECTED_VOTES. And both systems need to have the SAME cluster group number and cluster password.

Each system should have VOTES=1 and EXPECTED_VOTES=2, this is strongly required, if the systems share common storage (e.g. SCSI, FC, CI, DSSI). The each system could only boot, if it sees the other system.

If you do not like this and you know what you're doing, you could use EXPECTED_VOTES=1 on both nodes, but then you risk disk corruption, if you have shared storage.

Volker.

[Jan van den Ende]

ajaydec,

Bart is right.

In addition, are you sure that

>>>
show cluster shows following output on both the node:

| SYSTEMS | MEMBERS |
|â â â â â â â â -â â â â â â â â â â â â â â +â â â â â â â â â |
| NODE | SOFTWARE | STATUS |
|â â â â â â â â +â â â â â â â â â â â â â â +â â â â â â â â â |
| WM6 | VMS V8.3 | MEMBER |
| WM3 | VMS V8.3 | BRK_NEW |
+â â â â â â â â -â â â â â â â â â â â â â â -â â â â â â â â â +
<<<

is accurate???

The STATUS "BRK_NEW" indicates, that as far as the viewing node can know, the viewed node is configured but not yet booted into the cluster. And it would be quite a surprise to see that status from node WM3 about WM3...

- are the nodes connected to the same ethernet segment, or is any (switching) device between them?
- if yes to the previous, is SCS prococol ( which is NOT IP ) passed?
- do WM3 and WM6 both have (SYSGEN param) VAXCLUSTER as 2?
- do they share the same system disk, or each have its own?
- if 2 System disks, do they both agree on cluster number and password?

With answers to that, we may be able to guide you further.

Proost.

Have one on me.

jpe

[Hakan Zanderau ( Anders]

Another thing that can be a problem when setting up two nodes that have their own systemdisk is that both have the the defaultname ALPHASYS.

Make sure the systemdisks have unique names.

Hakan

[Hoff]

IP is not involved with clustering with any released version of OpenVMS; DECnet or IP can be used for cluster management and for downloading satellite images, but neither is used with the cluster communications and connectivity.

That you think IP is involved here implies that you're unfamiliar with the contents of the clustering manual. If you've not read that manual, go read it now. Seriously. Why? Clustering - configured incorrectly - risks your disk data; the disk data corruptions that can arise within a misconfigured cluster here are, well, complete.

[comarow]

Just the most common hunch,
Nisc_load pea0 Must equal 1
they must be connected no only on
the San, but also on the LAN.

[#8]

Hi All,

I have performed following steps to form a LAN Cluster b/w WM3 and WM6.


Steps performed for cluster configuration on WM3
------------------------------------------------


$ @sys$startup:cluster_config.com

Cluster Configuration Procedure
Executing on an Alpha System

DECnet-Plus is installed on this node.

DECnet is not running on this node.
LANACP LAN server process is running.
LANACP will be used for all satellite boot serving.

Cluster Configuration Procedure
CLUSTER_CONFIG_LAN Version V2.74
Executing on an Alpha System

DECnet-Plus is installed on this node.
Alpha satellites will use LANCP, not DECnet, for MOP downline loading.

Enter a "?" for help at any prompt. If you are familiar with
the execution of this procedure, you may want to mute extra notes
and explanations by invoking it with "@CLUSTER_CONFIG_LAN BRIEF".

This Alpha node is not currently a cluster member.

MAIN Menu

1. ADD WM3 to existing cluster, or form a new cluster.
2. MAKE a directory structure for a new root on a system disk.
3. DELETE a root from a system disk.
4. EXIT from this procedure.

Enter choice [4]: 1
Is the node to be a clustered node with a shared SCSI/FIBRE-CHANNEL bus (Y/N)? n
What is the node's SCS node name? WM3
Will the LAN be used for cluster communications (Y/N)? y
Enter this cluster's group number: 5
Enter this cluster's password:
Re-enter this cluster's password for verification:
Will WM3 be a boot server [Y]?
Verifying LAN adapters in LANACP database...
Updating LANACP LAN server process volatile and permanent databases...
Note: The LANACP LAN server process will be used by WM3 for boot
serving satellites. The following LAN devices have been found:
Verifying LAN adapters in LANACP database...

LAN TYPE ADAPTER NAME SERVICE STATUS
======== ============ ==============
Ethernet EWA0 ENABLED

Enter a value for WM3's ALLOCLASS parameter [0]:
Does this cluster contain a quorum disk [N]?

The EXPECTED_VOTES system parameter of members of a cluster indicates the
total number of votes present when all cluster members are booted, and is
used to determine the minimum number of votes (QUORUM) needed for cluster
operation.

EXPECTED_VOTES value for this cluster: 2

The use of a quorum disk is recommended for small clusters to maintain
cluster quorum if cluster availability with only a single cluster node is
a requirement.

For complete instructions, check the section on configuring a cluster
in the "OpenVMS Cluster Systems" manual.


WARNING: WM3 will be a voting cluster member. EXPECTED_VOTES for
this and every other cluster member should be adjusted at
a convenient time before a reboot. For complete instructions,
check the section on configuring a cluster in the "OpenVMS
Cluster Systems" manual.

Execute AUTOGEN to compute the SYSGEN parameters for your configuration
and reboot WM3 with the new parameters. This is necessary before
WM3 can become a cluster member.

Do you want to run AUTOGEN now [Y]? y



Steps performed for cluster configuration on WM6
------------------------------------------------


$ @sys$startup:cluster_config.com

Cluster Configuration Procedure
Executing on an Alpha System

DECnet-Plus is installed on this node.

DECnet is not running on this node.
LANACP LAN server process is running.
LANACP will be used for all satellite boot serving.

Cluster Configuration Procedure
CLUSTER_CONFIG_LAN Version V2.74
Executing on an Alpha System

DECnet-Plus is installed on this node.
Alpha satellites will use LANCP, not DECnet, for MOP downline loading.

NOTE: To ensure that this procedure executes in the correct context,
please invoke it from the system manager account.

Enter a "?" for help at any prompt. If you are familiar with
the execution of this procedure, you may want to mute extra notes
and explanations by invoking it with "@CLUSTER_CONFIG_LAN BRIEF".

This Alpha node is not currently a cluster member.

MAIN Menu

1. ADD WM6 to existing cluster, or form a new cluster.
2. MAKE a directory structure for a new root on a system disk.
3. DELETE a root from a system disk.
4. EXIT from this procedure.

Enter choice [4]: 1
Is the node to be a clustered node with a shared SCSI/FIBRE-CHANNEL bus (Y/N)? n
What is the node's SCS node name? WM6
Will the LAN be used for cluster communications (Y/N)? y
Enter this cluster's group number: 5
Enter this cluster's password:
Re-enter this cluster's password for verification:
Will WM6 be a boot server [Y]? y
Verifying LAN adapters in LANACP database...
Updating LANACP LAN server process volatile and permanent databases...
Note: The LANACP LAN server process will be used by WM6 for boot
serving satellites. The following LAN devices have been found:
Verifying LAN adapters in LANACP database...

LAN TYPE ADAPTER NAME SERVICE STATUS
======== ============ ==============
Ethernet EIA0 ENABLED
Ethernet EIB0 ENABLED

Enter a value for WM6's ALLOCLASS parameter [0]:
Does this cluster contain a quorum disk [N]?

The EXPECTED_VOTES system parameter of members of a cluster indicates the
total number of votes present when all cluster members are booted, and is
used to determine the minimum number of votes (QUORUM) needed for cluster
operation.

EXPECTED_VOTES value for this cluster: 2

The use of a quorum disk is recommended for small clusters to maintain
cluster quorum if cluster availability with only a single cluster node is
a requirement.

For complete instructions, check the section on configuring a cluster
in the "OpenVMS Cluster Systems" manual.


WARNING: WM6 will be a voting cluster member. EXPECTED_VOTES for
this and every other cluster member should be adjusted at
a convenient time before a reboot. For complete instructions,
check the section on configuring a cluster in the "OpenVMS
Cluster Systems" manual.

Execute AUTOGEN to compute the SYSGEN parameters for your configuration
and reboot WM6 with the new parameters. This is necessary before
WM6 can become a cluster member.

Do you want to run AUTOGEN now [Y]? y

After this I allow the AUTOGEN to run, after around 5-10 mins both the systems came up but they were not in Cluster.
Please note, I am able to setup up a LAN Cluster by following the above steps on other Alpha VMS systems.

Jan,

Sorry for creating a confusion.

"show cluster shows following output on both the node: .... " I mean to say both the nodes shows other node as BRK_NEW.

So show cluster on WM6 shows:

| SYSTEMS | MEMBERS |
|â â â â â â â â -â â â â â â â â â â â â â â +â â â â â â â â â |
| NODE | SOFTWARE | STATUS |
|â â â â â â â â +â â â â â â â â â â â â â â +â â â â â â â â â |
| WM6 | VMS V8.3 | MEMBER |
| WM3 | VMS V8.3 | BRK_NEW |
+â â â â â â â â -â â â â â â â â â â â â â â -â â â â â â â â â +

And show cluster on WM3 shows:

| SYSTEMS | MEMBERS |
|â â â â â â â â -â â â â â â â â â â â â â â +â â â â â â â â â |
| NODE | SOFTWARE | STATUS |
|â â â â â â â â +â â â â â â â â â â â â â â +â â â â â â â â â |
| WM3 | VMS V8.3 | MEMBER |
| WM6 | VMS V8.3 | BRK_NEW |
+â â â â â â â â -â â â â â â â â â â â â â â -â â â â â â â â â +

- are the nodes connected to the same ethernet segment, or is any (switching) device between them?
- if yes to the previous, is SCS prococol ( which is NOT IP ) passed?
- do WM3 and WM6 both have (SYSGEN param) VAXCLUSTER as 2? YES
- do they share the same system disk, or each have its own? Each have its own
- if 2 System disks, do they both agree on cluster number and password? YES

Hakan, both the systemdisk has different name.

Robert i tried rab(at)hp dot com but I got Delivery failure message. Could you please provide me your mail id so that i can send u a email.

Regards,
ajaydec

[Volker Halle]

ajaydec,

what's the value of SCSSYSTEMID on both nodes ? On each node: VOTES=1, EXPECTED_VOTES=2 - correct ?

Any messages on the console during boot about the 'other' node ?

Volker.

[Volker Halle]

ajaydec,

BRK_NEW - System has just booted, but has not yet joined the cluster and the connection manager has lost communication with it.

So the system 'see' each other via the LAN, but they won't allow the other system into the cluster.

Could you provide the console output from the boot of both systems ?

Volker.