June 1999
Callegra
Callware Technologies
8911 South Sandy Parkway
Sandy, UT 84070 USA
Phone: 801-984-1100
Fax: 801-984-1120
Web: www.callware.com
The telecommunications industry is experiencing a paradigm shift as companies start
moving away from closed, proprietary systems - which restrict their ability to choose and
integrate a variety of computer telephony (CT) applications - to open systems.
Historically, the lack of standards for inter-vendor cooperation has made it difficult for
CT software vendors to specialize in one particular product segment because they can't
count on other suppliers to fill niches or complement their product line.
Open standards present an opportunity to move beyond vendor interest toward an
initiative which is designed to benefit the entire industry. For example, imagine two
video game manufacturers, ABC and XYZ. Both companies make systems which use CDs for game
storage, but the CDs are not interchangeable. If someone gave you a game manufactured for
the ABC system, but you had an XYZ system, then you couldn't play the game. Your machine
wouldn't read what's on the CD.
However, if you could pop an ABC game into your XYZ system (or vice versa) and play it,
people who own either system would be happy because they could purchase either brand of
game to use in their machines. Additionally, the manufacturers could enjoy working
harmoniously with other vendors and competition would spur continual innovation in product
development.
Callware Technologies is helping to promote open standards-based telephony by
developing the first unified messaging application for CT servers, supporting Dialogic CT
Media server software and Microsoft Windows NT server. CT Media is an open systems-based
server software developed by Dialogic that fully embraces open standards for media
processing to interoperate with multi-vendor applications and hardware. CT Media supports
applications written to the ECTF S.100 API, and is being extended to support TAPI. By
expanding Callegra for Windows NT to include support for CT Media, Callware is increasing
the product's openness, scalability, and ability to interoperate with other
media-convergence technologies.
With Callegra on CT Media, users can run stand-alone unified messaging, add Callegra to
servers running multiple applications such as PBX or call center applications, and build
complete solutions that perform as an integrated system. Callegra's integration with CT
Media allows for sharing of data and call handling, as well as application-level features
such as Microsoft Exchange server integration to unified messaging, TTS, and Internet
telephony.
Callware's commitment to deliver unified messaging products based on open standards is
making it possible for solution providers to mix and match both hardware and software
applications to best meet the needs of their customers. The Callegra CT Media product has
been designed to complement other standards-based CT Media solutions, such as ACD, IVR,
and call accounting applications.
Callegra is an advanced messaging product for managing voice and fax messages that
integrates with e-mail packages such as Microsoft Outlook/Exchange, Lotus Notes, and
Novell GroupWise, or can operate as a stand-alone messaging platform. Callegra integrates
with e-mail, so users can see and respond to their voice and fax messages from their
e-mail box. Likewise, Callegra integrates with the Internet, so travelling users have
remote access to their messages. Users can also integrate multiple Callegra servers, using
the Internet as a transport medium for all types of messaging.
Callegra scales up to 96 ports, and supports Windows 95/98, and Windows NT-based
workstations and servers. It supports every major PBX/KSU vendor and with this release has
added support for several advanced digital switches, including those from popular
telephony system vendors such as Lucent, Nortel, Nitel, and NEC.
IP Launch Pad
Open Port Technology, Inc.
676 N. St. Clair Street
Chicago, IL 60611
P: 800-OPT-FAX-1
F: 312-664-3334
Web: www.openport.com
Any student of economics — for that matter,
anyone who has wondered about the contemporary fetish for the stuffed-doll-of-the-week
— is aware of the complex relationship between supply and demand. An excessive supply
of a particular commodity drives prices down, with the result that producers of that
commodity are forced to compete solely based on price: a good formula for going out of
business. On the other hand, excessive demand leads to frustrated consumers who begin to
consider other ways to get what they need: in the extreme, a good formula for civil
unrest.
The shifting balance between supply and demand is what drives both innovation and
profit in the open market, and nowhere is the need for this balance clearer than when
service providers try to make money on the Internet. Whereas traditional commerce requires
supply lines — trucks, trains, ships — to deliver products, the Internet
utilizes a less tangible medium known as bandwidth. Where traditional commerce delivers
mainly tangibles — books, foodstuffs, oil, etc. — the Internet delivers
information and applications. And where traditional commerce results in hard cash —
greenbacks, denarii, spondolees — Internet commerce results in soaring stock prices
and baffled economists.
Let’s step back for a moment. One of the biggest applications to date on the
Internet has been e-mail, and you will notice that, barring large file attachments,
Internet e-mail delivery happens rapidly and reliably. If e-mail delivery were as
unreliable as, say, a large file download or a real-time multiple-player game, it would
not be as popular as it is now.
Now consider the future of the Internet. You have various carriers and ISPs looking to
provide additional bandwidth to consumers — at this point, mainly businesses —
and charge them for it. But the trouble is that there are not currently that many
applications that require this bandwidth. Certainly, there is not the mainstream consumer
interest in paying for this additional bandwidth simply to download an occasional file
faster.
Enter Open Port Technology’s IP Launch Pad. Open Port is firmly convinced that the
way to interest consumers in additional bandwidth, and the way to get them to pay for that
additional bandwidth, is to provide applications. And the way to provide applications is
through a network-independent services creation platform.
With IP Lauch Pad, service providers, VARs, and LECs can launch and manage IP-based
services from a single, integrated platform. Pipes are pipes, but filled pipes are
products.
And IP Launch Pad is designed to allow service providers of any size to "get their
feet wet" with delivering services to a certain market, while retaining the option to
build-out their network to serve a larger market as their applications catch on. IP Launch
Pad provides complete hardware abstraction, so that third-party developers can stay
focused on developing solutions, rather than deciphering hardware and software
specifications. Also, the platform includes a Java-based Customer Management Administrator
(CMA) for network administration and troubleshooting.
IP Launch Pad supports standard industry protocols, including T.37, SNMP, IMAP4, LDAP,
and SMTP, and it supports both Solaris and Windows NT operating systems. Open Port also
offers an SDK for applications developers. System requirements for running IP Launch Pad
for NT are a 450MHz Pentium II, 64 MB RAM, 4 GB hard disk, and for Solaris are a 333 MHz
Ultra 10 with 2MB ECache, 128 MB RAM, and a 9 GB hard disk. Interested service providers
should contact Open Port for pricing information.
LoadBlaster 500
Hammer Technologies
205 Lowell Street
Wilmington, MA 01887
Phone: 800-HAMMER-IT
Fax: 978-988-0148
Email: hammer@hammer.com
Web: www.hammer.com
Hammer Technology's new LoadBlaster 500 is a CTI stress-tester tool designed for
quality assurance and manufacturing laboratories, or for any group that needs quick and
powerful low-level tests. The new design's main improvement is the use of a
"TestBuilder" GUI similar to those included with CTI graphical app-gen suites,
replacing the older Visual Basic and SuiteMaster editors found in Hammer's higher-end
developers' testing tools, although the VB editor is still available as an option.
A TMC Labs engineer recently visited Hammer's offices, and was given the opportunity to
see and test the LB500 in a hands-on environment. There wasn't enough time to conduct a
full-scale product review, but here are some initial impressions.
The LB500 is sold entirely as a turnkey system, based on industrial PCs from I-Bus
running Windows NT Workstation, or it can operate on the same platform as an existing
Hammer IT system. Natural MicroSystems provides the boards, with up to 12 spans per
chassis. Five chassis can be daisy-chained, and this 60-span configuration handles up to a
million calls per hour. Units can be rack-mounted, supporting T1, E1, ISDN (PRI) and SS7,
and multiple systems can be controlled with either a local controller or a LoadStation
remote master controller.
There's very little configuration involved for the end user - it just plugs in and
works after setting a few simple parameters - and, with just a couple hours of training,
we were able to understand the new testing model and create basic evaluation applications.
The documentation wasn't yet finished, but expect a quick-start guide, an administrator's
book, online help, and (best of all) an error code guide, both in print and in online
context, that converts cryptic messages into simple English.
TestBuilder's configuration option is for global settings. This option is more advanced
than Hammer's current BoardUtility option - it allows for system-wide settings of channel
protocols, phone books, multistage dialing, error handling, etc. Error handling (Figure 1)
is especially notable here. Every incident can take one of several options, including
hang-up and continue, stop script, stop test and associated channels, and freeze.
TestBuilder itself is designed for factory-floor technicians rather than R&D
groups; it's based on flow charts and actions with English names instead of programming
code. Actions and properties follow Windows conventions - right single-click for this,
left double-click for that - and their descriptions are intuitive. Sample actions include
PlayPrompt; ReleaseCall; Send Digits; PlaceCall; Fax/ReadySend; Fax/SendFax; SendTone; and
Pause. There's also an extremely useful ConfirmPath option, which sends raw data straight
out and back. This test/pause function is useful for keeping a connection active that
would normally time out; it would appear as a horizontal line, compared to the diagonal
lines in Figure 2. The up- and down-arrow buttons let users switch an action's place
within a test.
Also include is the Library Manager, which houses pre-made basic testing scripts. These
scripts can be implemented as part of a larger solution or used as the foundation for
customized scripts. Another new option lets users save scripts in personal folders, which
in turn are accessible from any PC on the network. This feature was conspicuously missing
from older Hammer systems.
TestBuilder's Scheduler option can handle an infinite number of tests. It offers
options for maximum testing time, staggered channel starts (random, user-defined, or
automatic) and maximum active connections. Users can also force all channels to
synchronize before a test begins, and tests can run based on a predefined number of loops
or a predefined length of time.
The Monitor feature hasn't changed much in appearance, but it is highly functional.
Monitor offers call summaries (attempts, completions, and failures), a system monitor
similar to the older VB editor, and a graphical call-flow status that shows active legs
and on-deck legs in different colors, line widths, etc.
TestBuilder Reports are also quite useful and revamped from older versions. In one
view, users can see a summary of all systems connected to one controller, with call
attempts and completions. Or, users can view a specific system, specific channels with
call details, or error reports that specify the system, channel, test, error ID, reason,
and action taken. It's even possible to view a test by specific time intervals - for
example, if you've let a test run over a weekend, you can come in on Monday and view
reports from Saturday, 4:56 p.m. to Sunday, 1:03 a.m. This is an invaluable debugging
tool.
Another new feature is the remote API, which will be in future releases. This API will
allow remote control of the system from a non-Hammer workstation, with the basic abilities
to start tests, stop tests, and retrieve data.
There are also plug-in options available, which add functionality to the configuration
and action tabs, and to the monitor, reports and canned scripts. Plug-ins will include
voice-over-IP loads, Fax, SS7, and ISDN testing.
Tests are configured using an interface that seems like a cross between Pong and an
app-gen . In this multistage example, an IVR is waiting for calls and receives one from
side A. The IVR plays its "send me a digit" prompt, which side A hears and
replies to with the appropriate action. The IVR verifies the digit and asks side A to now
send a digit to side B. Side A complies, and side B receives the digit, prompts side A for
a tone, receives and verifies it, and then releases the call. Side A hears, understands
the release, and disconnects.
NETWORK RESIDENT SERVER V. 3.0
ObjectSwitch Corporation
900 Larkspur Landing
Suite 270
Larkspur, CA 94939
P: 415-925-3471
F: 415-925-3499
Web: www.objectswitch.com
ObjectSwitch says version 3 of its network resident telecommunications server software
provides an enterprise server package that contains true fault tolerance, extreme
scalability and speed, open architecture, and the flexibility of hot swappable software
applications.
The company says the new ObjectSwitch server software, which supports the Rational Rose
object modeling environment, can scale to millions of users, handle tens of thousands of
transactions per second, and is distributable across multiple systems. It runs on the Sun
Solaris, HP-UX, and Windows NT server platforms.
ObjectSwitch 3.0's network resident server technology replaces complex, hand-coded
infrastructure, eliminating the need for middleware, and allowing developers to focus on
business logic while reducing time-to-market and development costs.
Pre-built core adapters and interfaces in ObjectSwitch 3.0 can handle Oracle, Sybase,
Informix, DB/2, CORBA, SNMP, and TCAP/SS7 systems. Custom modules and applications include
GSM encoding, automated service provisioning (ASP), call detail record (CDR) management,
LHS (Edifact), and Tuxedo. Supported configurations include TCP/IP, X.25, HTTP, UDP/IP,
POP3, FTP, 327X and serial interfaces.
With its self-generating code, ObjectSwitch 3.0 seamlessly integrates telecom,
Internet, and data networks, bringing the world of object modeling to super-scale,
memory-resident database servers. Applications can be switched on the fly, without any
system downtime, all the objects picked up and flowing through new interfaces.
All this means development time can be greatly truncated. Grover P. Righter, vice
president of marketing for ObjectSwitch, said the company delivered an integrated server
system to Caribbean Wireless in five weeks. "And the specification was for 18
months," Righter said.
Transparent code-generated recovery, queues that tolerate external system failure, and
software engine auto recovery give ObjectSwitch 3.0 true fault tolerance for
mission-critical enterprise applications.
Robert Shavell, an analyst with Data Monitor in New York, said the new ObjectSwitch 3.0
release represents a quantum leap in performance and reliability of enterprise server
software.
"It's far and away one of the highest performing transaction servers,"
Shavell said. "The hot swappable feature, few if any competing products have. And the
high throughput, the number of transactions it can process, is pretty much unprecedented
for big-time telco switching products."
While saying ObjectSwitch's market is more specialized than some of its competitors -
with customers like U.S. West, France Telecom Mobile, and Nextel - "what it does for
telco operators is a very big deal," Shavell said.
Anne Thomas, a market analyst with the Patricia Seybold Group in Boston, was equally
impressed with ObjectSwitch 3.0.
"There are two huge features. The model-based development environment that allows
you to build an enterprise-class application without writing code, and the fact that the
engine supports massive scalability, recoverability, and reliability."
Thomas said ObjectSwitch successfully borrowed its architecture techniques from
telephone switching systems, transplanting them to the application server arena.
"When you pick up a phone, you expect a dial tone. And they've taken all that
solid core technology to make it work for applications systems."
ObjectSwitch products include a telecom integration server, a transaction application
server, a design center IDE, an object model auditor, ASP and CDR pre-built services, plus
various custom adapters and interface modules.
|
TMCnet LOGIN
Webinars
