 |
Description  |
|
|
TECHNICAL FIELD
This invention relates generally to a method of software management, and
more specifically to the process of maintaining forward and backward
compatibility for computer systems utilizing various software releases
containing different database information.
BACKGROUND OF THE INVENTION
In computer systems that employ distributed networks where database
information is frequently transferred, there exists the need to maintain
both forward and backwards compatibility among the various software
releases generated during a systems lifetime. Examples of computer systems
utilizing distributed networks are Local Area Networks (LANs) and Wide
Area Networks (WANs). In these computing environments, forward
compatibility describes the relationship between a progression of software
releases, whereby installation of a new software release does not
compromise the fitness, form, or function achieved under previous software
versions. In essence, the new software performs like the old, despite the
inclusion of additional features. Backwards compatibility is achieved when
the reinstallation of a previous software version does not render the
system inoperable.
This same need arises in computer systems that employ redundant database
information which is either compared or exchanged. One such example is a
telephony computer designed to perform the switching of customer calls. An
example is the Motorola Electronic Mobile Exchange (EMX) family of
cellular switches. These include the Motorola EMX 100, 250, 500, and 2500
switch families which support cellular radiotelephone services in several
major metropolitan markets. The interested party may receive full
system/hardware descriptions on such devices by contacting Motorola's
Cellular Publishing Services at 1501 W. Shure Drive Shure Drive, Arlington
Heights, Ill. 60004 and requesting Instruction Manuals 68P8105196E-99E,
68P81052E-54E and 68P81056E for the EMX 100-500 Family of Switches or
Instruction Manuel 68P09201A07-A for the EMX2500 electronic switch all of
which are incorporated herein by reference.
Simply stated, a telephony computer is nothing more than a large switch
that employs sophisticated software capable of managing and directing
multiple customer calls per second. In this effort, it is necessary to
develop a comprehensive customer database capable of reporting the various
telephone subscribers, their individual accounts, and various service
features. Presently, most of the customer database information is of a
static nature; not subject to frequent change and therefore suitable for
storage in backup form. With the expansion of customized telephone
services, however, a growing portion of the database information is
dynamic.
Unlike static, dynamic information is customer generated, subject to
certain alteration, and therefore totally unsuitable for hard copy
storage. Typical examples of dynamic information include the programmable
automatic redial, call forwarding, busy transfer, no answer transfer, or
voice mailbox telephone options. While this personalized information is
not maintained in hard copy form, it is extremely valuable to the average
system subscriber and therefore must be jealously safeguarded during
database transfers if quality phone service is to be provided.
In this effort to insure quality service, EMX computers employ a redundant
or secondary switch that is a duplication of the primary switch, and
capable of continuing service if the primary is ever disabled. During
normal operation, the primary and secondary customer databases are
frequently compared. These subscriber audits are performed in order to
assure primary and secondary database equivalence. In addition, it is
quite common for telephony computers to perform subscriber file transfers,
wherein the subscriber information residing in one computer is transferred
to a different computer. Another frequently performed operation is
subscriber feature preservation. This is the process whereby dynamic
database information is updated and transferred to the secondary switch
the instant the primary becomes disabled. In each of these operations, the
existence of forward and backwards compatibility is imperative in order to
assure the proper handling and transfer of the appropriate database
records. This is especially true for the transfer of dynamic information
which is constantly changing and for which there is no hard copy backup.
Forward and backwards compatibility is normally achieved on an individual
software release basis. It will be appreciated by those skilled in the art
that a typical software update may include the correction of an old
software version or the addition of new system features. Each enhancement,
however, requires altering the database records supporting the capture of
the obsolete or newly acquired information. Such changes present a
considerable challenge to the programmer attempting to implement forward
and backwards compatibility, because communication between differing
database structures may result in the loss of important information. In
the telephone business, lost information represents an intolerable breach
in the quality of service.
For example, assume an original software release version 1.0 is supported
by customer database #1, which contains several records each having
elements A, B, and C. The updated release of this software is version 2.0
which is supported by customer database #2. Database #2 also contains
several records, however, these records contain elements A, B, and D.
During the forward transfer of database #1 information into database #2,
it is understood that the software associated with database #2 will
disregard the information found in element C. Likewise, during the
backwards transfer of database #2 information into database #1, the
software associated with database #1 will not comprehend the information
found in element D. Ultimately, this unrecognized data must be discarded
by the software. Of note, the greater the difference between the subject
databases and the more software releases to be encountered, the more
complicated system software must be in order handle these inconsistencies
and maintain compatibility.
Previous approaches suggest downloading the entire existing database,
undesired records and all, and relying upon complex conversion programs
capable of converting the original database to a format capable of being
passed to the new database. While these programs are entirely capable of
discarding obsolete records or ignoring the presence of new records,
compatibility will remain a serious problem as the number of operating
software releases escalates. Each new release must maintain compatibility
with its predecessors, while the predecessors must often be updated in
order to successfully communicate with newer releases. This represents a
formidable task which is virtually impossible of being fault tolerant.
In order to avoid the needless waste associated with sending unused
information across transmission lines of limited capacity, the prior art
also suggests manually enter the appropriate database information. Due to
the sheer size of some databases, however, manual entry depicts a labor
intensive process which is extremely prone to operator error, and
incapable of handling the real time demands imposed by dynamic database
information.
It would therefore be extremely advantageous to provide a simplistic method
for transferring database information between various processors having
different database information or structures, while maintaining forward
and backwards compatibility.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a flow chart diagram of the steps performed by a digital computer
during database transfers in accordance with the present invention;
FIG. 2 is a block diagram example of a slave process database language
description;
FIG. 3 is a block diagram example of a master process database language
description;
FIG. 4 is a block diagram example of the working language; and
FIG. 5 is an example of the compacted database information described by the
working language of FIG. 4 in working language format.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
The present invention has general applicability within computer based
devices and systems employing various software releases which typically
contain different data base information. According to the preferred
embodiment, these devices are processors utilized within Motorola's family
of EMX telephony computers. As such, they may contain different versions
of system software during a processor upgrade and/or installation
procedure. In order to maintain both forward and backwards compatibility
for such systems during software upgrades, the data base transfer program
depicted in FIG. 1 is executed.
FIG. 1 is a flow chart diagram of the steps performed by such a telephony
computer during a database transfer in accordance with the present
invention. According to this method, an initiating processor, hereinafter
referred to as the master processor requests an at least a second
processor, hereinafter referred to as the slave processor to provide a
description of the slave's database language. Upon receipt of this
description, the master compares the slave's database language description
to that of the master's. Looking specifically to the areas of comparison,
the master collects the items common to either languages. The information
that is foreign to both languages is ignored. In this way the master
develops a new database language description which facilitates
communication between the master and the slave based upon their mutuality.
As previously mentioned, this common description is called the working
language.
The working language is the intersection between the master and the slave
database language descriptions. As such, it contains the set of elements
common to both languages. It therefore represents an area of mutual
compatibility between various database configurations.
Once the working language is developed, its description is sent back to the
slave process. From this point, both processes are capable of requesting
the exchange of database information. In operation, the master and the
slave employ the working language in order to process the actual database
information exchanged. In this context, processing includes, but is not
limited to, compacting the database information into the condensed working
language format, or expanding compacted database information into its
expanded form for subsequent handling.
This disclosed method of transferring database information resolves the
problem of maintaining forward and backwards compatibility during software
release updates by creating a separate database language description
compatible with both releases. In this way database translations are
achieved in a method independent of complex conversion programs.
When a database information exchange request is received, the responding
(slave) process extracts information from its database and places it into
the working language format. Since this format is understood by both
master and slave, communication is effectuated, despite the differences
between these software releases.
Upon receipt of the requested information, the requesting (master) process
will process the requested database information. Accordingly, the master
will extract the requested information from the working language format
and place it in the appropriate database records of its current
configuration. At this point the master is capable of performing the
previously mentioned subscriber audits, subscriber file transfers, or
subscriber feature preservations, independent of the existing software
releases.
Unlike the prior art, the information foreign to both processes is ignored
during the development of the working language. This step avoids the
needless waste associated with sending unused information across
transmission lines, and storing that information prior to discard. As a
results, the development of the working language greatly enhances overall
information throughput due to these timely savings.
For illustrative purposes FIG. 2 is a block diagram example of a slave
process database language description. As such, it depicts the type of
information found in a slave database. While the description does not
contain the actual values stored in the slave database, it nonetheless
describes the format these actual values will follow.
The first line of the database description is called the language
structure. This section performs a preamble function and serves to
describes the body of the language. The variables the language structure
uses in its description are: Language count, Entry size, Compacted length,
and Database type. By definition: Language count gives the number of
entries to be found in the body of the language; Entry size gives the size
of each entry; Compacted length is not used in the slave database language
description; and Database type identifies the type of database this
language describes.
According to the example in FIG. 2, the slave database language description
has only four entries. Additionally, each entry has three variables.
Compacted length is only used by the working language, therefore a zero
fills that variable. Compacted length will be further explained in
connection with the description of the working language. Finally, the
Database type carries a designation of one. The Database type variable is
provided in order to accommodate system flexibility. In the instance that
several different database formats are employed, this variable allows the
system to differentiate the various formats during working language
development.
Immediately following the language structure appears the body of the
language. The body consists of numerous entries that describe the various
features of the slave process. In this effort, each entry employs
variables that describe the features format. The variables the entries use
in their description are: Feature ID; Feature type, and Feature Length.
As previously mentioned, each entry corresponds to an individual process
feature. For identification purposes, each entry is therefore given a
unique identifier called a Feature ID. After designation, that Feature ID
will remain dedicated to that specific feature across all future software
releases.
For example, assume the slave process is capable of supporting the
telephone option, call forwarding. Within the slave's database language
description, there will be an entry describing this particular feature.
For identification purposes, assume the entry describing call forwarding
is given the Feature ID 1. In each new software release utilizing call
forwarding, there will be an entry identified as Feature ID 1, which is
identical in all respects to the slave's entry describing the call
forwarding option. For those releases not utilizing call forwarding, there
will be no entry having Feature ID 1.
Each entry will also have a Feature type variable. Feature type describes
the form of the information utilized in this entry. There are three
Feature types: byte type, bit type, and end bit type. If the feature is
described in byte (s) of information then it is byte type, and is
identified by the literal value of 0. If the feature is described in
individual bit(s) of information, then it is bit type and identified by
the literal value of 1. End bit type is only used by the working language,
and will therefore be explained in connection with the description of the
working language.
Finally, each entry will have a Feature length variable. This variable
designates the length of the database feature being described. This length
will be in bytes of information if the feature is of the byte type, or in
bits if the feature is of bit type.
Returning to FIG. 2, it will be appreciated that the first entry in the
slave's database description describes a feature identified as Feature ID
1. Feature ID 1 comprises five bytes of information. The next entry
describes Feature ID 2, which describes a feature three bits in length.
The next entry is Feature ID 3, which describes a feature consisting of a
single bit of information. Finally, the last entry describes a feature
designated as Feature ID 5. Feature ID 5 describes a feature that consists
of a single byte of information.
In comparison, FIG. 3 is a block diagram example of a master process
database language description. At a glance, the language structure reveals
that this database description contains six entries which utilize the same
variables as the slave database description. This is recognizable because
both the slave and the master have identical entry size and database type
designations. In addition, the master process employs all the features
that were available under the slave process configuration. The difference
is that the master process utilizes two additional features; Feature ID's
6 and 7 which are both a single byte in length.
According to the present invention, when either process desires to initiate
communications with the other, the requesting process is deemed the
master. The master therefore requests the database language description of
the responding or slave process, compares the slave's description of the
database language body to that of its own, and develops a working
language.
FIG. 4 is the block diagram example of the working language developed from
the comparison of FIG. 2 and FIG. 3. The working language is merely the
intersection between the master and slave database languages, and
therefore contains only the Feature ID's common to both. Because the
master utilized all the features present in the slave process, FIG. 4 is
almost identical to the description in FIG. 2. Of course this will not
always be the case, but for illustrative purposes, it will be appreciated
that the intersection between the body of the slave's database language
description and that of the master's is represented in FIG. 4. In this
example, the only areas of difference are the previously mentioned end bit
and compacted length variables.
The end bit type variable is a pseudo feature type variable specific to the
working language, thus it will only appear in the body of the working
language, and only as a Feature type variable entry. In essence, the end
bit literal informs requesting software when the last bit of the bit type
information stored in a particular byte has been reached. The software is
thereby provided an indication when to point to the next byte of
information. According to FIG. 4 Feature ID 3 is the last bit type
information common to both the master and slave. Therefore, Feature ID 3
has the end bit literal value of 2 as the feature type designation.
The only remaining difference is the compacted length variable. As
previously discussed, compacted length is a field variable used only by
the working language. This variable informs the requesting software of the
length of compacted database information, i.e., the width of the
information in the working language format.
As previously mentioned, all database information is compacted prior to
exchange in order to save space. For example, Feature ID's 2 and 3 require
a total of 4 bits of representation. These two Feature ID's will therefore
be placed in a single byte prior to transmission to the requesting
process. As a results, the compacted database information will comprise
five bytes of data for Feature ID 1, a sixth byte for data pertaining to
Feature ID's 2 and 3 data, and a seventh byte for Feature ID 5 data. The
compacted database information described by the working language is 7
bytes in length. Accordingly, the compacted length variable for the
working language is 7. FIG. 5 is an example of the compacted database
information described by the working language of FIG. 4.
In review, a method of maintaining backward and forwards compatibility
between various software releases having different database information
has been described. According to the present invention, various software
releases seeking to communicate, compare database structures and develop a
common language. This common or working language facilitates subsequent
database transfers, independent of each process software release. It will
nonetheless be appreciated that this disclosed method will also work
effectively wherein the various software releases contain the same
information or structure.
While only particular embodiments of the invention have been shown and
described herein, it will be obvious that additional modifications may be
made without departing from the spirit of this invention. It is therefore
not intended that this invention be limited, except as indicated by the
appended claims.
* * * * *
|
|
 |
|
 |
Description |
|
