Naming in Db2 data sharing

By maintaining a consistent naming convention, you can simplify the process of naming the many items in a Db2 data sharing environment. You also simplify the management of name changes, as you add new members or move existing members to different systems in the future.

About this task

You assign to names to entities during the process of installing Db2 and enabling data sharing. Some parts of a data sharing system require you to specify a valid name. Other names are generated during the installation process.

After installation, you cannot change some names, such as the group name and member names. To change names of data sharing members you must quiesce existing members and add replacement members with the new name. For more information about that process, see Renaming data sharing members.

Assign names to both IRLM and data sharing groups, and to members within a group. One recommendation is to make names and prefixes unique within the Parallel Sysplex®. Although this uniqueness is not required for all names, it helps you avoid problems with identifying and moving entities among z/OS systems in the Parallel Sysplex.

If you are enabling the originating member of the group from an existing Db2 subsystem, you can build a naming scheme around existing names to reduce the disruption to existing applications. However, before enabling data sharing, you might want to change some names to lay the foundation for a solid naming convention. For information about the process for changing the name of the originating member, see Renaming the originating Db2 subsystem for a new data sharing group.

Another name to choose carefully is the catalog alias for the group. It is very difficult to change that name. To change the catalog alias for the group, you must bring the entire group down and perform the single-system procedure for every member of the group.

Procedure

Consider the following suggestions when naming various Db2 entities:

Name subsystems first and use the subsystem names as the basis for creating other names.

Subsystem names are limited to four characters, making them the shortest names.

Four-character names are also needed for the following entities:

Db2 group name; one per data sharing group
Member name; one for each Db2 subsystem in the group
IRLM subsystem name; one for each member of the group

One possible naming convention is a Parallel Sysplex-wide approach to naming. This convention assigns subsystem names of the form ctmg where:

c: Denotes a particular collection of logically related applications or subsystems.
t: Denotes a type of resource, such as B for Db2 or J for IRLM.
m: Denotes a particular member within a data sharing group or an IRLM group. This identifier also associates a subsystem with its z/OS® system for recovery when you use automatic restart.
g: Denotes a particular Db2 group.

This information uses A as the group identifier. Therefore, the naming scheme has a subsystem name of DB1A for the first member of the data sharing group. The second member is DB2A, and so on. The following table shows how the member identifier relates to a particular z/OS system and how the group identifier associates members across a Parallel Sysplex system. In each column are the names that are associated with a particular z/OS recovery group (z/OS2, for example). In each row are the names that are associated with the primary components of the data sharing group, such as Db2 subsystems and IRLM subsystems.

Table 1. Subsystem names in a Parallel Sysplex
z/OS1	z/OS2	z/OS3	z/OS4
DA11 (CICS®)	DA21 (CICS)	DA31 (CICS)	DA41 (CICS)
DB1A (Db2)	DB2A (Db2)	DB3A (Db2)	DB4A (Db2)
DJ1A (IRLM)	DJ2A (IRLM)	DJ3A (IRLM)	DJ4A (IRLM)
DD11 (DBCTL)	DD21 (DBCTL)	DD31 (DBCTL)	DD41 (DBCTL)

If you enable an existing Db2 subsystem for data sharing, and existing applications already access that Db2 subsystem, consider using the name of the existing Db2 subsystem as the group attachment name. This strategy allows existing applications continue to use the data sharing member. It does not require any changes to those applications in terms of specifying a new subsystem name.
Use the same name for the Db2 group name, the Db2 location name, and the Db2 integrated catalog alias.
A data sharing group uses the same catalog alias name. This catalog alias is used as the high-level qualifier for the Db2 directory (DSNDB01), catalog (DSNDB06), default database (DSNDB04), and work file database VSAM data sets.
Use the same names as the member name and Db2 subsystem name for each member.
Accept the default command prefix for each member.

Example of names for a Db2 data sharing environment

Table 2. Db2 data sharing naming example
Entity	Name
Db2 group name	DSNDB0A
ICF catalog alias	DSNDB0A
Group attachment name	DB0A
Subgroup attachment name	SBG1
Catalog and directory database names	DSNDB0A.DSNDB01 and DSNDB0A.DSNDB06
Application defaults module	DSNHDECP
Target libraries	DSNDB0A.SDSNCLST DSNDB0A.SDSNLINK DSNDB0A.SDSNLOAD DSNDB0A.SDSNEXIT
Network Address	DOMAIN DSNDB0A.svl.ibm.com AND PORT 446
Group DRDA Port	446
Group Secure DRDA Port	448
Location name	STLEC1
Generic LU name	USIBMSY.SYEC1GLU
Parallel Sysplex domain name	DSNDB0A.svl.ibm.com
Group IPv4 Address	1.1.1.1
Group IPv6 Address	1.1.1.1.1.1.1.1.1
Member name	DB#A
Db2 subsystem name	DB#A
Location alias name	ALIAS1
Member LU name	LUDB2LA
Member domain name	DSNDB0A.svl.ibm.com
Network address	DOMAIN DSNDB0A.svl.ibm.com AND PORT 5001
Resync Port	5001
Member IPv4 Address	2.2.2.2
Member IPv6 Address	2.2.2.2.2.2.2.2.2
Command prefix	-DB#A
Work file database	WRKDB#A
Load module for subsystem parameters	DSNZP0#A
Member BSDS names	DSNDB0A.DB#A.BSDS01 DSNDB0A.DB#A.BSDS02
Active log data set prefixes	DSNDB0A.DB#A.LOGCOPY1 DSNDB0A.DB#A.LOGCOPY2
Archive log data set prefixes	DSNDB0A.DB#A.ARC1 DSNDB0A.DB#A.ARC2
SCA and group buffer pool connection names	DB2_DB#A
Procedure names	DB#AMSTR DB#ADBM1 DB#ADIST DB#ASPAS
XCF member name	DB0A
IRLM group name	DXRDB0A
IRLM subsystem name	DJ#A
IRLM procedure name	DB#AIRLM
IRLM member ID	Number 1-12 corresponding to #
IRLM XCF member name	DXRDB0A$$DJ#A12
Lock structure connection name	DXRDB0A$$DJ#A12
Cache structure (group buffer pools)	DSNDB0A_GBP0
Lock structure	DSNDB0A_LOCK1
List structure (shared communications area)	DSNDB0A_SCA

What to do next

Consider a convention for names for distributed processing. Those naming conventions are probably be part of a much broader convention. See your network administrator for more information about choosing names for distributed access.