 |
Description  |
|
|
CROSS REFERENCE TO RELATED APPLICATIONS
U.S. patent application, Ser. No. 07/261,861, entitled "Method to Provide
Concurrent Execution of Distributed Application Programs by a Host
Computer, and an Intelligent Work Station on an SNA Network" filed
10/24/88 and assigned to the Assignee of the present invention, is
directed to a method for executing LU 6.2 conversations for distributed
application programs which involves creating a pool of virtual machines at
the host which are maintained in a run ready idle state until assigned by
a Virtual Machine Pool Manager, to a request from a terminal executing a
first part of a distributed application program, for a conversation with
the second part of the distributed application program stored at the host
system.
U.S. patent application, Ser. No. 07/250,598, entitled "Method to Manage
Concurrent Execution of a Distributed Application Program by a Host
Computer and a Large Plurality of Intelligent Work Stations of an SNA
Network" filed 9/29/88 now U.S. Pat. No. 4,940,254 issued Aug. 14, 1990
and is assigned to the Assignee of the present invention, is directed to a
method for providing improved communications between distributed portions
of an application program that is being executed on a network
simultaneously by a relatively large number of terminals, in which a pool
of virtual machines is created at the host and each machine is primed
(initialized) with the host resident portion of the program. When a
conversation request is received at the host and assigned to a virtual
machine in the pool, the machine is ready to accept the request and begin
processing immediately. The pool manager monitors the use of the pool
relative to anticipated demand and adjusts the size accordingly in
accordance with an established algorithm.
FIELD OF INVENTION
This invention relates in general to LU 6.2 type inter-program
communication methods in SNA data processing networks in which terminals
are not preassigned in the system with specific end users and in
particular to a method for permitting an end user to establish a current
terminal address which is valid only while the terminal is active or until
terminated by the user.
BACKGROUND ART
The prior art discloses a variety of computer networks. The IBM System
Journal, Volume 22, Number 4, 1983 includes a series of articles devoted
to a review of the IBM System Network Architecture (SNA). On page 345 of
that publication a network is defined as "a configuration of terminals,
controllers, and processors and the links that connect them. When such a
configuration supports user applications involving data processing and
information exchange and conforms to the specifications of the System
Network Architecture it is called an SNA network. Essentially SNA defines
logical entities that are related to the physical entities in a network
and specifies the rules for interactions among these logical entities.
The logical entities of an SNA network include network addressable units
and the path control network that connects them. Network addressable units
communicate with one another using logical connections called "sessions."
The three types of Network Addressable Units (NAUs) are the Logical Unit
(LU), the Physical Unit (PU), and the System Services Control Point (SSCP)
which are defined as follows;
Logical Unit (LU). An LU is a port through which end users may access the
SNA network. An end user uses an LU to communicate with another end user
and to request services of a System Services Control Point (SSCP).
Physical Unit (PU). A PU is a component that manages the resources of a
node in cooperation with an SSCP.
System Services Control Point (SSCP). This is a focal point for
configuration management, problem determination and directory services for
end users. SSCPs may have sessions with LUs and PUs. When such a session
occurs, the LU or PU is in the domain of the SSCP. In addition to sessions
with LUs and PUs, SSCPs may also communicate with each other to coordinate
the initiation and the termination of sessions between Logical Units and
in different domains."
From the hardware standpoint, a simple network comprises a host system
having a processing unit and a plurality of remote terminals that are
assigned to individual users. The remote terminals are selectively
connectable to the host system through one or more communication links.
These links may comprise merely a coaxial cable, a dedicated telephone
line, or in some cases, a satellite communication link.
The host processing unit most always has an operating system which supports
the creation of a large number of virtual machines or the functional
equivalents, each of which is assigned, on request, to an end user. A
virtual machine processes tasks for the assigned end user, by time sharing
the host processor hardware of the host system. Some hosts systems may
include more than one hardware processor so that true simultaneous
processing occurs at the host since a plurality of processors are running
in parallel. More often, there is merely one hardware processor that
"concurrently" runs data processing tasks for the virtual machines by a
time sharing technique. This is transparent to the end users at the
terminals.
In prior art networking systems, a user of the system is assigned a USERID
by the system administrator at the time the user is first authorized to
use the system. The user is also assigned a Password at that time, which
is required to be presented to the system each time the User "Logs On" to
the system. The system administrator has entered the user's USERID, and
Password into the system. If the information that is provided by the Log
On process does not match the information entered into the system by the
administrator, the user in not allowed access to the system. If the
information is correct then the system allows access. As part of the Log
On process the User is permanently assigned the exclusive use of a Virtual
Machine by the system until a Log Off process is completed. The name given
to the assigned Virtual machine is the USERID of the user that is Logging
On. The system address of the terminal being used by the user for the Log
On process is also transmitted to the assigned virtual machine and stored
as part of the Log On process. Any subsequent communications intended for
the user, require only the user's USERID as part of the massage address.
The message is transferred to and processed by the assigned virtual
machine and sent to the user at the terminal address stored by the
dedicated virtual machine at the time of Log On.
This method of assigning a user's USERID as the name of the virtual machine
that has been assigned to that user and storing the user's current
terminal address in the assigned virtual machine, avoids the address
conversion problem of converting USERIDs to terminal addresses in order to
deliver a message addressed to a specified USERlD . With this prior art
approach the system only requires the USERID of the intended recipient in
order to deliver a message. When a message is received for a user it is
stored and a check is made to determine if the message can be delivered
immediately. If the user is on line at the time the system receives the
message, the system will indicate this after scanning a relative short
list of virtual machines that were dedicated to specific users at Log On,
and proceed accordingly to deliver the message by causing the one
dedicated virtual machine having the name USERID to transmit the message
to the user's current terminal address stored by that dedicated virtual
machine. Otherwise, as part of the Log On process, the system checks the
list of undeliverable messages for messages addressed to the user and
advises the newly signed on user that there is a waiting message stored in
the system.
Two general types of terminals are employed in data processing networks.
The first is referred to as a "dumb terminal" in that it comprises merely
a keyboard and a display device and little or no processing capability
other than that required to make a connection with the host system. The
second type of terminal is referred to as an Intelligent Work Station
(IWS) and is provided with its own processor unit, Operating System and
supporting peripheral devices. The terms IWS and Personal Computer (PC)
are often used interchangeably. With the ready availability of PCs having
very attractive price performance characteristics, most new networks are
implemented with IWS type terminals and many of the older networks are
being modified with the replacement of dumb terminals with IWS type
terminals.
Providing each end user on the network with its own processing capability
relieves the host CPU from doing many of the data processing tasks that
were previously done at the host. The nature of the tasks that are
processed by the host CPU therefore has changed and more sophisticated
applications such as electronic mail and electronic calendaring are now
implemented on the network under the control of the host system. Both of
these applications involve what is referred to as distributed application
programs, in that one part of the application program is resident on the
host system and another is resident on the IWS terminal.
Many of the current data processing networks are designed in accordance
with the IBM SNA architecture which was first described in 1974. Since
then various new functions and services have been added. As suggested
earlier, SNA networks can be viewed as a plurality of nodes interconnected
by data links. At each of these nodes, path control elements send
information packets, referred to as Path Information Units (PIUs), between
resource managers called Logical Units. The logical connections of the
paths are called a session. A transport network for data is therefore
defined by the path control elements and the data link control elements.
Nodes can be connected by a plurality of links and comprise a plurality of
LUs. Various types of LUs sessions and protocols have been established
within the framework of the SNA architecture. There are three general
classes of sessions. The first class is unspecified by SNA. The second
class involves terminals and the third involves program to program
communication. For example LU 6 provides SNA defined inter-program
communication protocols which avoids the limitations of terminal LU types
such as LU 2 and LU 7. LU 6.2 is referred to as Advanced Program to
Program Communication or APPC protocols.
Logical Units are more than message ports. LUs provide operating system
services such as program to program communication involving one or more
local programs. Each application program views the LUs as a local
operating system and the network of loosely coupled LUs connected by
sessions as a distributed operating system.
The LU allocates a plurality of resources to its programs, which are
dependent on the particular hardware and its configuration. Some of the
resources that are made available are remote while others are local, i.e.,
associated with the same LU as the application program. The sessions are
considered local resources at each LU, but are shared between particular
LUs.
The control function of an LU is resource allocation. Programs ask one for
access to a resource. Sessions which carry messages between LUs or
programs running on LUs are considered shared resources. A session is
divided so that a plurality of conversations are run serially.
Two LUs connected by a session have a shared responsibility in allocating
sessions to application programs for use as "conversations." The
application programs are therefore sometimes referred to as "transaction
programs."
The successful connection between LUs occurs as a result of a common set of
protocols which function first to activate a session between two LUs and
second to facilitate the exchange of message data.
The SNA format and protocol reference manual designated SC30-3112,
published by the IBM Corporation describes SNA by describing, for example,
with programming language declarations, the format of messages that flow
between network entities and the programs that generate, manipulate,
translate, send and return messages.
The SNA transaction program reference manual for LU 6.2 referred to as
GC30-3084, published by the IBM Corporation defines the verbs that
describe the functions provided by the implementing products.
Intelligent work stations that are connected to a SNA type network and
employ an LU 6.2 protocol to process an application program that is
distributed between the IWS and the host system operate efficiently so
long as the operating system of the IWS does not run more than one
application concurrently at the terminal. However, if the IWS is operating
under an operating system such as OS/2, which allows an IWS such an IBM
PS/2 personal computer to run concurrent application programs which are
distributed, the advantage of concurrent operation on the PS/2 is lost.
The advantage is lost because at the host, the separate transactions which
are run concurrently at the terminal become serialized. The serialization
of the transaction occurs because the host creates only one virtual
machine that is permanently associated with the user ID and the specific
terminal as long as the session is active.
In order to avoid the serialization at the host, the second application
being run at the terminal has to be run with a different user ID in order
to have a separate virtual machine established at the host that will be
dedicated solely to the second application.
The invention described in the cross-referenced application Ser. No.
07/261,861 is directed to a method to permit two or more distributed
application programs that are being run concurrently on one intelligent
work station of a data processing network to be executed on separate
virtual machines created by the host system to prevent the applications
from becoming serialized at the host and to allow each to be run
concurrently with the other on both the host and the terminal.
With the method of the cross-referenced application, the host system
creates a plurality of virtual machines (VMs) that are brought to a run
ready state prior to and in anticipation of being assigned to a
distributed application program for processing a task which has been
defined in said distributed application program, part of which is resident
on the host system and the companion part of which is resident on one of
the IWS end user terminals. The pool of run ready VM machines are
preferably created automatically at the time that the host system is
initialized under the control of a pool manager, which is a program
resident on the host system, whose other main function is to assign an
idle VM machine from the pool in response to an end user request that
identifies a distributed application program, a previously assigned
Logical Unit name and a USERID. The VM is assigned only for a period of
time required to complete one LU 6.2 conversation. At the end of the
conversation the VM machine is returned to the pool for subsequent
assignment to another, possibly different, application program and user.
The method allows two distributed application programs being executed
concurrently on the IWS to run concurrently on the host in two separate
virtual machines even though the conversation requests have the same
USERID.
While the above system improves the processing of distributed application
programs, it requires a new method to determine the current terminal
location of an on-line user. It will be recalled that in the prior art
method, the dedicated virtual machine that was named the USERID of the
current user, was used to store the current terminal address of the user.
This dedicated machine has effectively been eliminated and replaced by a
pool of virtual machines which are not permanently associated with either
one user or one terminal address. Since virtual machines from the pool are
assigned dynamically to process relatively short LU 6.2 type conversations
between the host and the terminals, and then returned to the pool, it is
not practical to use the prior art method of determining the current
terminal address of an active user by naming an assigned virtual machine
from the pool with the USERID of the user and storing the user's current
terminal address in this virtual machine.
The method of the present invention, allows the current terminal address of
a specific user in an SNA system employing a pool of virtual machines
organized according to the teaching of the above Cross-referenced
applications to be determined only after the intelligent workstation
terminal user has decided that the host system or other system users may
communicate to the user's intelligent workstation.
SUMMARY OF THE INVENTION
In accordance with the method of the present invention, a specific user of
an intelligent work station such as an IBM PS/2 type personal computer,
running a multi-tasking type operating system such as the IBM OS/2
operating system and which is connected to an SNA network having a host
processing system employing LU 6.2 program to program communication
protocols decides when the host system or other users on the system may
deliver messages addressed to the USERID of the specific user employing
those protocols. The method establishes a distributed Sign On application
program in the system with one part of the program resident on each IWS
and the other part resident at the host processor. When the IWS user
decides that communication with the host processor and other users is in
order, the Sign On distributed application program is evoked on the IWS.
The Sign On program displays a screen to the user prompting for the user's
USERID and the password that was assigned when the user was authorized to
use the system. The workstation then issues an LU 6.2 ALLOCATE verb
requesting a conversation with the counterpart program that is specified
in the ALLOCATE. The Part B of the Sign On program, which has the name
"BEGIN", is executed on the host processor to process the conversation
being requested by the ALLOCATE. The Begin program calls a program which
is named VMPMID, that is a component of the Virtual Machine Pool Manager
(VMPM) program. A Identification Control Block Entry (ICBE) is created by
this component which functions to store USERID and Terminal ID information
in respective fields along with a pointer to a following ICBE if one
exists. A DEALLOCATE verb is then issued by the BEGIN counter part program
and the conversation is terminated with a message to the user indicating a
successful conclusion. The method further stores the chain of ICBEs in
storage at the host processor which is accessible to any program that is
run on a virtual machine that is a member of the pool of virtual machines.
The method also requires that any distributed application program that
originates at the host processor and intends to issue an allocate verb to
initiate a conversation involving a specific user, first scan the ICBE
chain to determine if the user is accepting messages, and if so the
current system address where these messages should be sent in accordance
with LU 6.2 protocols.
The user is allowed to revoke the permission initially granted by evoking
the Sign On program again and selecting the end option. An ALLOCATE verb
is then issued by the terminal, requesting a conversation with a
counterpart application program called "END" which is another component
program of the VMPM program. The ALLOCATE transmits sufficient information
to allow the END component program to locate the appropriate ICBE in the
chain of ICBEs and cause it to be deleted. The END program then issues a
DEALLOCATE back to the terminal, which display a message to the user
indicating a successful completion of the permission revoking task.
It is therefore an object of the present invention to provide an improved
method for executing distributed applications in a data processing
network.
A another object of the present invention is to provide an improved method
for processing distributed application programs in an SNA type data
processing network in which a pool of virtual machines is established at
the host processors.
A further object of the present invention is to provide an improved method
for processing distributed application programs in an SNA type network
employing LU 6.2 protocols, whereby an end user determines when messages
addressed to the user's USERID from other system users may be delivered to
the user.
A still further object of the present invention is to provide an improved
method for identifying the current terminal address of an end user in an
SNA network which employs LU6.2 protocols in which the host processor does
not create a virtual machine that is dedicated to one user at one
terminal.
A still further object of the present invention is to provide a method for
identifying the current address of an intelligent workstation being
operated by a specified user of an SNA network employing LU 6.2 protocols
in which the specified user is not required to execute prior art log on
procedures in order to use the workstation.
Objects and advantages other than those mentioned above will become
apparent from the following description when read in connection with the
drawing.
DESCRIPTION OF THE DRAWING
FIG. 1 is a schematic design of a data processing network.
FIG. 2 is a schematic representation of one of the IWS terminals shown in
FIG. 1.
FIG. 3 illustrates the organization of the various layers of programming
that are involved in the SNA network of FIG. 1.
FIGS. 4A and 4B show the relationships between parts of a distributed
application program and network programs.
FIG. 5 is a schematic representation of the pool of run ready virtual
machines that are created at the host processor.
FIG. 6 illustrates the details of the Virtual Machine Pool Data Structure
that is employed the by the Pool Manager in managing the pool of virtual
machines shown in FIG. 5.
FIG. 7 is a flow chart setting forth the steps involved in creating the
pool, of virtual machines shown in FIG. 5.
FIGS. 8a and 8b are flow charts setting forth the steps involved by the
Pool Manager in executing a distributed application program in accordance
with the new method.
FIG. 9 is a flow chart setting forth the steps involved in advising the
host processor of the current terminal address of a system user.
FIG. 10 illustrates the Identification Control Block Entry (ICBE) data
structure that is created to store the current terminal address of an
online system user.
FIG. 11 is a flow chart setting forth the steps involved when the host
processor attempts to initiate an LU 6.2 conversation with a system user.
FIG. 12 is a flow chart setting forth the steps involved when an online
user decides not to receive any further conversations with the host
processor.
DESCRIPTION OF THE PREFERRED EMBODIMENT
FIG. 1 illustrates an information handling system comprising an SNA network
20 of interactive type terminals or Intelligent Work Stations (IWS) 21 of
the type shown in detail in FIG. 2. As described, the network includes a
plurality of terminals 21 which are interconnected to a host central
processing system 23. As shown in FIG. 1, host 23 in turn is connected by
communication link 24 to a host processing system 25, which also connects
to another SNA network 26 of interactive terminals 21. Functionally, the
system operates to allow each terminal or end user to communicate with the
host and to one or more other terminals or users using established SNA
communication protocols so that the various serially connected
communication links are transparent to the users.
The host system includes a host processing unit which may by way of example
be an IBM 370 system. A virtual machine type operating system, such as the
IBM VM Operating Systems, is assumed in the description of the preferred
embodiment.
It should be assumed that the SNA network shown in FIG. 1 supports two
distributed applications referred to as "MAIL" and "CALENDAR" which are
available to each terminal user. The MAIL application program allows a
user at one terminal to generate a document such as a letter and send that
letter to one or more other users at a designated nodes on the network.
The sender can store the document in the host system at some logically
central system location. Each addressee of the letter has the capability
of retrieving that document at a later time by also using the MAIL
application program from his terminal. The CALENDAR application functions
to maintain an electronic calendar for each terminal user. The CALENDAR
application, for example, allows one end user to view other end users'
calendars prior to scheduling a meeting in order to determine free periods
of those persons being invited to the meeting. Such systems are well known
in the art and are currently an extensive commercial use. Since the
general organization and operation of such distributed applications is
well known, only those details that are necessary for an understanding of
the method of processing data in distributed application programs of the
present invention will be described.
It should therefore be assumed in the following description that each
workstation on the network is an Intelligent Work Station such as an IBM
PS 2 personal computing system employing a multitasking operating system
such as the IBM OS/2 Operating System. It may be further assumed that
conventional SNA services to support Logical Unit type LU 6.2 for
distributed applications are provided by the system. The terminal shown in
FIG. 1 may therefore process two distributed application programs such as
MAIL and CALENDAR concurrently.
FIG. 2 illustrates the functional components of one of the interactive type
data processing terminals 21, shown in FIG. 1. The terminal comprises a
processing unit 31, which includes a microprocessor block 32, which is,
for example, an Intel 80386 micro-processor, a semi-conductor memory 33, a
control block 34 which functions to control input-output operations in
addition to the interaction between the microprocessor block 32 and the
memory unit 33.
The terminal further includes a group of convention peripheral units
including a display device 36, keyboard 37, printer 38, a storage unit 39,
and modem 40. Since the details of the above described functional blocks
form no part of the present invention and can be found in the prior art,
only brief functional description of each block is set forth along with
the description of their interaction, sufficient to provide a person of
ordinary skill in the art with the basis of understanding applicant's
improved method of processing distributed application programs
concurrently.
Processing unit 31 corresponds, for example, to the system unit of an IBM
personal computer such as the IBM PS/2 model 80 system. Unit 31 is
provided with an operating system program which may be the IBM
multi-tasking OS/2 operating system which is normally employed to run the
PS/2 model 80. The operating system program is stored in memory 33 along
with the application programs that the user has selected to run. When the
system supports a distributed application program such as MAIL or
CALENDAR, only one part, e.g., part A of the distributed application
program is stored at the terminal while the other part, part B, is stored
at the host system. Depending on the capacity of memory 33 and the size of
the application programs, portions of these programs as-needed may be
transferred to memory 33 from the disk storage unit 39 which may include,
for example, a 40 megabyte hard disk drive and a diskette drive. The basic
function of storage unit 39 is to store programs and data that are
employed by the system and which may readily be transferred to the memory
unit 33 when needed. The function of the diskette drive is to provide a
removable storage function of entering programs and data into the system
and a vehicle for storing data in a form that is readily transportable for
use on other terminals or systems.
Display 36 and keyboard 37 together provide for the interactive nature of
the terminal, in that in normal operation the interpretation that the
system gives to a specific keystroke by the operator depends, in
substantially all situations, on what is being displayed to the operator
at that point in time.
In some situations the operator, by entering commands into the system,
cause the system to perform a certain function. In other situations, the
system requests the entry of certain data generally by displaying a prompt
type of menu/message screen. The depth of the interaction between the
operator and the system varies by the type of operating system and the
application program, but is a necessary characteristic of terminals on
which the method of the present invention may be employed.
The terminal shown in FIG. 2 further includes a printer 38, which functions
to provide hard copy output of data. Lastly, the modem 40 functions to
transfer data from the terminal 21 of FIG. 2, to a host system through one
or more SNA communication links.
FIG. 3 shows the various layers of programming that are employed in an SNA
type network. The SNA programming environment is generally considered to
consist of seven layers as shown. The top layer as shown is the End User
layer and consists of the end user programs. The second layer is called
the NAU Services. These services include, for example, presentation
services, terminal services and formatting data for specific applications.
The third layer is referred to as Data Flow Control. Its function is to
maintain send/receive modes and perform high level error correction. The
fourth layer is the data Transmission Control layer. Its function involves
such things as encryption and decryption plus session level pacing. The
fifth layer is the Path Control which does routing, segmenting data units
and virtual route pacing. The Data Link layer is the sixth layer. It
functions to provide link level addressing, sequencing and error control.
The seventh and last layer is the Physical layer which defines for example
the pin assignments on connectors for the various signals.
APPC defines the NAU services, Data Flow Control and Transmission Control.
As explained on page 306 of the previously referenced IBM Systems Journal,
the method of defining the LU 6.2 conversation functions, is in terms of
programming-language-like statements called verbs. Documentation with
verbs which are completely defined by the procedural logic that generates
session flows, provides significantly greater precision than English
prose. FIG. 4A shows how the verbs define the interaction between
transaction programs, i.e., Part A or Part B of the distributed
application, and Logical Units for conversation resources. A set of verbs
is referred to as a protocol boundary rather than as an application
program interface.
As shown in FIG. 4A, the presentation services component interprets verbs
and can be thought of as including a subroutine for each verb. The LU
resource manager does allocation of conversation resources and assignment
of conversations to the sessions, keeping queues of free sessions and
pending allocation requests. Its equivalent component in products also
allocates local resources in products specific ways. The function of the
following LU 6.2 verbs is set forth on page 307 of the previously
mentioned IBM System Journal. The 6.2 verbs discussed are one, SEND.sub.--
DATA, RECEIVE.sub.-- AND.sub.-- WAIT, PREPARE.sub.-- TO.sub.-- RECEIVE,
FLUSH, REQUEST.sub.-- TO.sub.-- SEND, SEND.sub.-- ERROR, CONFIRM, ALLOCATE
AND DEALLOCATE.
The ALLOCATE verb initiates new activity at another LU by building a
conversation to a named partner program. The named partner is placed in
execution and given addressability to the conversation that started it.
The ALLOCATE verb carries several parameters including the following.
1. LU.sub.-- NAME. This is the name of the LU at which the partner program
is located.
2. TPN. TPN is the Transaction Program Name of the partner program with
which the conversation is desired.
3. MODE.sub.-- NAME. MODE.sub.-- NAME specifies the type of transportation
service that the conversation is to provide. For example, a SECURE, a
BULK, or a LOW.sub.-- DELAY conversation can be requested. The LU uses a
session with the appropriate MODE.sub.-- NAME to carry the conversation.
The target of the conversation is a newly created process or task, which
means that the distributed processing in the network at any instant of
time, consists of a number of independent distributed transactions, each
of which consists of two or more transaction programs connected by a
conversation. The DEALLOCATE verb ends the conversation. In as much as
each partner may issue DEALLOCATE, a conversation varies from a single
short message to many exchanges of long or short messages. A conversation
could continue indefinitely, terminated only be a failure of a Logical
Unit or by the session that carries it. Transaction programs are not ended
by DEALLOCATE, but continue until they terminate their own execution, end
abnormally or are terminated by control operator action.
Both network application programs and service transaction programs use the
execution services provided by Logical Units. Service transaction programs
run on Logical Units in the same way as other transaction programs. They
interact with the human operator or they may run as a pure programmed
operator. Many service transaction programs effect only the local Logical
Unit. An example is a command to display the current set of active
transaction programs.
Other control transactions, especially those that relate to sessions, can
effect other Logical Units as well as applications at other Logical Units.
For example, a local command to prematurely terminate a transaction that
is using a conversation causes the conversation to be ended abnormally, a
state change that must be transmitted to the partner Logical Unit for
presentation to the transaction program that is sharing the conversation.
Or a decision to activate one or more of the sessions shared by the two
LUs may be made by one LU operator but must be communicated to the other
Logical Unit. Advanced program to program communication for SNA includes
several control operator verbs that provide LU to LU control and
coordination, especially for activation and deactivation of sessions. When
a distributed service transaction program starts at one LU, it creates a
conversation to a partner transaction program in a partner LU. The two
transaction programs then cooperate to preform the desired control
activity.
The IBM VM host operating system includes a component referred to as
APPC/VTAM Services (AVS) which is responsible for the APPC protocol
boundary support in the Operating System. AVS defines one more LU 6.2
Logical Units to IBM Virtual Telecommunications Access Method (VTAM). VTAM
is the IBM host computer component that manages the communications layer
between the host and the various terminals of the network. AVS acts as a
bridge for APPC communications to virtual machines within the operating
system. For example, when an APPC ALLOCATE verb is received that
originated from outside the VM operating system, VTAM will determine if
there is a Logical Unit active that corresponds to the LU name specified
in the ALLOCATE. AVS will have previously told VTAM that it will handle
all traffic for particular LU names. VTAM will find that AVS has defined
an LU that corresponds to the LU name in the ALLOCATE verb and pass the
ALLOCATE verb to AVS.
There is additional information supplied with the ALLOCATE verb that is
used in this process. Included in | | |
