DFSMSrmm Primer and Conversion to CA-1 Updates
Note: This is publication is now archived. For reference only.
Published 26 January 2005
Rate and comment
Authors: Mary Lovelace
With DFSMSrmm z/OS V1.6 or later you can share a single DFSMSrmm control data set between multiple single systems or multiple sysplexes without the need to have a shared DASD environment. This Technote introduces you to the RMMplex.
With DFSMSrmm z/OS V1.6 or later you can share a single DFSMSrmm control data set between multiple single systems or multiple sysplexes without the need to have a shared DASD environment. This new function was announced with DFSMSrmm z/OS V1.6 and is call Client/Server Support. To use this new function you must have TCP/IP installed on all systems sharing a single DFSMSrmm control data set, and all systems must be at the z/OS V1.6 level.
An RMMplex is one or more MVS systems each running a DFSMSrmm subsystem sharing a control data set. An RMMplex can optionally include one or more DFSMSrmm subsystems as servers, one or more client subsystems, in addition to standard DFSMSrmm subsystems. The server subsystems and standard subsystems have direct access to the DFSMSrmm control data set and share it. The client systems have no direct access to the DFSMSrmm control data set, but share the control data set via the server. All systems that share a control data set in this way are part of the same RMMplex.
A “server” in DFSMSrmm terminology is a z/OS system with DFSMSrmm installed and active. This DFSMSrmm has direct access to the DFSMSrmm control data set and support requests coming from other systems via TCP/IP.
A “client” in DFSMSrmm terminology is a z/OS system with DFSMSrmm installed and active. This DFSMSrmm has no direct access to the DFSMSrmm control data set, and sends all requests to the DFSMSrmm control data set via TCP/IP to the server specified in EDGRMMxx PARMLIB member.
RMMplex parameters in EDGRMMxx
Following are the EDGRMMxx PARMLIB parameters in support of the RMMplex.
This specifies the identifier of the control data set that must be used on this system. Specify a value one to eight alphanumeric characters long. When you start DFSMSrmm, the CDSID ID is compared to the ID in the control data set control record. If the IDs match, DFSMSrmm startup continues. If the control data set does not have an ID, DFSMSrmm creates the ID in the control record from the CDSID. If the IDs do not match, DFSMSrmm startup fails and DFSMSrmm issues a message to the operator to select another PARMLIB member. If you do not specify a value for CDSID, you cannot use a control data set that already has an ID in its control record. See the section “Creating or Updating the Control Data Set Control Record” in the manual DFSMSrmm Implementation and Customization Guide, SC26-7405, for information about how the DFSMSrmm EDGUTIL utility sets the control data set ID.
This specifies the type of system you want to set up. CLIENT is mutually exclusive with SERVER. If neither client nor server are specified, DFSMSrmm starts as a standard system.
The servername can either be an IP address, a fully qualified domain name, or a server host name. DFSMSrmm uses the domain name system (DNS) to resolve a domain name or a host name into an IP address. SERVERNAME is a required operand when you specify CLIENT. Servername can be 63 alphanumeric characters, period (.), and hyphen (-). The host name can be a maximum of 63 characters. The host name must contain one or more tokens separated by a period. Each token must be larger than one character. The first character in each token must start with a letter. The remaining characters in each token can be a letter, number, or hyphen. For example, CLIENT(SERVERNAME(RMMPLX1.MAINZ.IB M.COM) PORT(1950)) tells DFSMSrmm to start as a client without direct DASD access and to share the tape inventory in access by the RMM server with the host name RMMPLX1 using network IP protocol port 1950.
Use this operand to specify the port number to be used for IP communication. The PORT operand is required. Specify a value from 1 to 65535. Port numbers 1 to 1023 are reserved. Also, the client port number and server port number must match for the systems to communicate.
This specifies the type of system you want to set up. SERVER is mutually exclusive with CLIENT. Neither SERVER nor CLIENT must be specified when DFSMSrmm is used as a standard system.
Use this operand to specify the port number to be used for IP communication. The PORT operand is required. Specify a value from 1 to 65535. Port numbers
1 to 1023 are reserved. The port number must be the same for the client system and the server system to establish a network connection.
Use this operand to specify how many DFSMSrmm tasks should be available on the server to handle socket connections from client systems. DFSMSrmm
uses this number to determine how many tasks are to be started for processing all client requests on this server. Specify a value from 1 to 999.
The number of local and server tasks you can use and still successfully start DFSMSrmm is limited by the size of the private region above and below 16 MB.
To start with more tasks, you will need a larger REGION size.
This material has not been submitted to any formal IBM test and is published AS IS. It has not been the subject of rigorous review. IBM assumes no responsibility for its accuracy or completeness. The use of this information or the implementation of any of these techniques is a client responsibility and depends upon the client's ability to evaluate and integrate them into the client's operational environment.
Follow IBM Redbooks
Follow IBM Redbooks