OSI/MHS Configuration and Management Manual

Managing Your OSI/MHS Subsystem
OSI/MHS Configuration and Management Manual—424827-003
7-3
Group Databases
Group Databases
Each GI, LO, MR, MS, and RS group has a PDU store database associated with it.
The MS group has an additional database associated with it: the SQL database. The
SQL database has its own catalog, table, and indexes. For a full discussion of
managing SQL databases, see the NonStop SQL Installation and Management
Manual.
An association exists between the message store PDU store database and the SQL
database. TMF keeps the backups of these two databases consistent. Otherwise,
when you recover to a known point, you could find messages referred to in the
mailboxes that are not in the PDU stores, or messages in the PDU stores that are not
in the mailboxes. TMF also provides audit trails relating to the PDU store and SQL
databases.
Monitoring the MHS Objects
In managing the OSI/MHS subsystem, you are actually monitoring three types of
objects. You monitor the objects that represent the executable portion of the
subsystem (the CLASS, GROUP, and PROCESS objects), you monitor the objects that
represent the information flow through the subsystem (the APPL, ROUTE, and MTA
objects), and you monitor the objects that represent the grouping of users within the
subsystem (DLIST, DLISTMEMBER, CUG, and CUGMEMBER). This section
describes how to manage these three categories of objects. For complete information
about SCF commands that apply to MHS objects, see the OSI/MHS SCF Reference
Manual.
When you have hundreds of objects in your subsystem, keeping track of everything
can be complex. One way to make it easier is to use naming conventions that help
you recognize specific objects and group related objects. When naming conventions
are meaningful, you can use wild-card characters effectively (as described in
Section 3, Planning Your OSI/MHS Subsystem). This will help you to manage your
subsystem more effectively.
Managing Groups
The group is the basic unit of configuration and recovery in the OSI/MHS subsystem.
You start a group and stop a group. Each group has a database associated with it. If a
failure happens to any process within a group, the entire group fails and the entire
group recovers. Therefore, you manage the executable portion of the system by
monitoring groups. At a general level, you can examine all the groups in a class. At a
specific level, you can examine all the processes in a group. This section describes
some of the SCF commands you can use to monitor groups in the subsystem.
Note. These examples are based on the configurations in Appendix A, Examples of
Configuration Files. The exact output you get when you try these commands will vary
according to the state of the processes at the time you execute the commands.