OSI/MHS Configuration and Management Manual
Sizing and Tuning Your OSI/MHS Subsystem
OSI/MHS Configuration and Management Manual—424827-003
8-4
Number and Distribution of Users
•
The MTA attribute REM-MSG-OVERHEAD specifies how long it takes a specific 
adjacent MTA to process the transmission, not including the time it takes to transfer 
the message on the line.
If you set MAX-MSG-XFER-TIME to 0 and specify values for REM-MSG-THRUPUT 
and REM-MSG-OVERHEAD, OSI/MHS sets a timer for each message, based on the 
message size and the values you specified, as follows:
(message size/REM-MSG-THRUPUT) + REM-MSG-OVERHEAD
Because the timer value is adjusted for each message, using these MTA attributes is 
potentially much more efficient than using MAX-MSG-XFER-TIME, especially in error 
situations.
Number and Distribution of Users
The resources required to support your user community depend on several factors, 
including:
•
The number of users (human or machine)
•
Their proximity to or association with existing MTAs
•
Their means of access to the system—for instance, X.25 or asynchronous 
connection, gateway or message store
•
Any naming requirements or constraints on users
These factors determine which OSI/MHS components and underlying software are 
needed to support the users. They also determine which O/R name types are most 
efficient for routing, including where and how to define distribution lists.
Number and Distribution of Messages
To plan the number of OSI/MHS groups of different types and the sizing of the 
OSI/MHS databases, you should know:
•
How many messages the system handles on average
•
What proportion of messages arrive from remote user agents, adjacent MTAs, and 
gateways
•
What proportions of messages are intended for relay, message store, and gateway 
delivery
•
How many recipients the average message has (because there is a separate 
routing action for each recipient)
Traffic Patterns
In addition to the number of messages expected on average, it is important to 
understand traffic fluctuations such as the number of messages that will arrive in peak 










