Previous topic |
Next topic |
Contents |
Contact z/OS |
Library |
PDF
Planning for distributed mode z/OS MVS Planning: Operations SA23-1390-00 |
|||||||||||||||||||
There are things that you need to consider before
migrating to distributed mode.
Note: The following discussion regarding
console attributes applies only to MCS, SMCS, and subsystem consoles;
it does not apply to EMCS consoles (except as noted).
In distributed mode, changes to the attributes
of a console persist only while the console is active. When a console
is deactivated, its attributes revert to their original values. The
original values for consoles in distributed mode are established either
when the console is first defined or when the sysplex enters distributed
mode with the SETCON MODE=DISTRIBUTED command. More specifically:
There are a few exceptions to the rule. The following
console attributes persist when a console is deactivated:
If you want to change the original attributes
of a console, you must redefine the console.
An alternate method for changing the original
attributes of an MCS, subsystem, or SMCS console is:
Note: This is potentially disruptive to your operations and
should only be used if you have not exploited the ability to define
more than 99 MCS, subsystem, or SMCS consoles. See Migration and fallback considerations.
In distributed mode, you cannot alter the attributes
of inactive consoles (including EMCS consoles) with the following
exceptions:
In distributed mode, you cannot alter the type of a console. The console name retains the console type until the console is deleted using IEARELCN or IEARELEC. In distributed mode, the health checker will only report on inactive MCS and SMCS consoles that are defined on the system on which the health checker is running. In distributed mode, console IDs for MCS, SMCS and subsystem consoles no longer have to be in the range of 1-99. Therefore, the use of console names and the CNZCONV service are recommended when trying to find information pertaining to consoles. If you have any local program that examines the console UCME array, you must use the CNZMXURF service to access the array and your program must obtain the local and CMS locks before invoking the CNZMXURF service. To obtain the CMS lock, your program must be authorized. Your programming must also be able to tolerate console IDs that are not the same as a console’s array index. Also, the UCME array will contain only information about MCS, SMCS and subsystem consoles that are active on this system. The tracking facility will identify users of the CNZMXURF service that are not holding the CMS lock. IBM® recommends using the CNZCONV service for obtaining information about consoles or parsing the response to the DISPLAY CONSOLE command. In distributed mode, if a console is already active
on another system and you attempt to issue certain VARY commands from
this system, you will see different messages from what you receive
in shared mode. For example, an MCS console named STEVE is both defined
on SYS1 and SYS2 in the sysplex. The console STEVE is already active
on SYS1. The table below lists the commands issued from SYS2 and the
messages in response when the console services support is in distributed
mode and shared mode.
|
Copyright IBM Corporation 1990, 2014
|