How to get others nodes' information in cluster

[olive_wide]

Two nodes cluster.
When the two nodes all are active,at one node ,how to get another's information?
I want to get tcpip
information ,like "hostname" , "ipaddress",etc.
Best regards
Olive

[Peter Quodling]

Explore the SYSMAN utility.

particular set env/node

q

[olive_wide]

Peter,
Can you tell me detailedly?
How to do.
olive

[Robert Gezelter]

Olive,

The underlying assumption is that you have access to a management-level account on the cluster.

To use SYSMAN in this situation, you would do the following:

$ MCR SYSMAN
SYSMAN>set environment/cluster
SYSMAN>do show network
SYSMAN>exit

SYSMAN will relay to "show network" command to each member of the cluster. There are limitations to this technique, but they do not apply in the case of the SHOW NETWORK command.

- Bob Gezelter, http://www.rlgsc.com

[Uwe Zessin]

You can also collect the output, e.g.:

SYSMAN> do show network/output= device:[dir]SHONET_'F$GETSYI("NODENAME")'.LIS

Or you run your own data collection routine:

SYSMAN> do @device:[dir]SHONET.COM

[comarow]

You have other options.

Show sys/node=othernode
then you can do any command in sysman
as has been well pointed out

mcr sysman
set env/cluster
do tcpip show version
do tcpip show interface
or whatever.

Then you have the show cluster/cont command
you can issue commands like

add hw_type
and find out what kind of cluster it is.
(use help to see all the commands)



I have to wonder why if you have a two node cluster you can't simply get the information? Is there something strange going on?

Bob

[olive_wide]

What will I do?
Maybe everybody want to know.
Because I find a way to achieve "BACKUP".
For example,two nodes cluster,one is TEST, another is BU.
If all the applying processes in server base on TCPIP service.It seems easy to achieve"BACKUP".
I make a program in startup_vms.com, when the system startup,for example TEST,there is a menu:
This node is TEST, which system you want to startup?
1.TEST
2.BU
then you can select 1 or 2.
If select 1, startup normal.
If select 2, runing:
$ DEFINE/SYSTEM/EXEC UCX$CONFIGURATION -
$ [dir]:UCX$CONFIGURATION.DAT;1
$DEFINE/SYSTEM/EXEC UCX$HOST -
$[dir]:UCX$HOST.DAT;1
$DEFINE/SYSTEM/EXEC UCX$NETWORK -
$[dir]:UCX$NETWORK.DAT;1
$DEFINE/SYSTEM/EXEC UCX$PROXY -
$[dir]:UCX$PROXY.DAT;1
$DEFINE/SYSTEM/EXEC UCX$ROUTE -
$[dir]:UCX$ROUTE;1
$DEFINE/SYSTEM/EXEC UCX$SERVICE -
$[dir]:UCX$SERVICE.DAT;1
This director [dir] is differet from sys$system ,it includes 6 files tcpip*.dat. These 6 files are tcpip config of BU!
after I define these 6 files, and startup.The system will startup as BU! Can you understang what I want to do?

Now, I want to get tcpip information of another node.Because I want to add something in my program.It is a "judgement", I don't know if it is correct word, because my poor english.
After I define the tcpip.dat I can startup BU as TEST, and startup TEST as BU. But when I startup BU as TEST, the TEST will NOT startup as TEST! if you select TEST, it will show error,it is what i want to do.Only to get another node's tcpip information(tcpip must have been startup) and decide local node can startup.

This is my idea, welcome to discuss.
Olive.

[Uwe Zessin]

That doesn't work, I am afraid. The TCP/IP databases are usually shared, so the configuration is tied to the SCS nodename.

If the system hardware is the same, then I would just boot the system into the other system root - else, I would create another system root with the required configuration data that you can boot from.

[olive_wide]

U,
It does work! :)
I didn't copy the tcpip file from BU server.
In TEST NODE:
I delete TEST's tcpip config, and change the tcpip config to BU,include proxy,route,service,etc. Then, copy this data to a [dir], and then reconfig the tcpip back to TEST.
I have test the program ,it does work well.
I haven't tried to create a new root for another node. Did you try it? Does it work well? Detail step?
Thanks a lot
Olive

[olive_wide]

U,
It does work! :)
I didn't copy the tcpip file from BU server.
In TEST NODE:
I delete TEST's tcpip config, and change the tcpip config to BU,include proxy,route,service,etc. Then, copy this data to a [dir], and then reconfig the tcpip back to TEST.so the database is good.can use in local node.
I have test the program ,it does work well.
I haven't tried to create a new root for another node. Did you try it? Does it work well? Detail step?
Thanks a lot
Olive