Technical White Paper-I
Delivering Managed Network
Services
Prepared By
Hughes Escorts Communication Limited
Plot No. 1, Sector 18, Electronic City,
Gurgoan 122 015
India
Document No. HECL/MNS/NC/001 Ver 1.0
Tuesday, May 07, 2002
Ver 1.0 Hughes Proprietary Page 1 of 1
Revisions History
S No. Description Owner Version
Number
1 Base created Neeraj C 1.0
Table of Contents
1.1. Network management: - a historical prospective ...................................................................... 3
1.2. Managed Network Services: An Overview................................................................................. 4
1.3. MNS Center Technical Architecture .......................................................................................... 6
1.4. Infrastructure Details................................................................................................................... 7
1.5. Network Management.................................................................................................................. 8
1.6. HECL Service Management Strategy....................................................................................... 16
1.7. Service Information Portal ........................................................................................................ 19
1.8. Application Management........................................................................................................... 21
1.9. System and Operations Management ....................................................................................... 22
Contact Information
For further information visit
Wide Web site at:
http://www.hughesindia.com
Email : - krish@hughes-ecomm.com
Conventions
S. No. Acronym What it means
1 MNS Managed Network Services
2 HECL Hughes Escorts Communication Limited
3 OV Open View
4
5
Ver 1.0 Hughes Proprietary Page 2 of 2
1.1. Network management: - a historical prospective
The early 1980s saw tremendous expansion in the area of network
deployment. As companies realized the cost benefits and productivity
gains created by network technology, they began to add networks and
expand existing networks almost as rapidly as new network
technologies and products were introduced.
By the mid-1980s, certain companies were experiencing growing pains
from deploying many different (and sometimes incompatible) network
technologies.
The problems associated with network expansion affect both day-to-
day network operation management and strategic network growth
planning. Each new network technology requires its own set of
experts. In the early 1980s, the staffing requirements alone for
managing large, heterogeneous networks created a crisis for many
organizations. An urgent need arose for automated network
management (including what is typically called network
capacity planning) integrated across diverse environments.
Basic Network Management
Management Entity
Network Management System
Servers
SNMP
Agent Proxy
Database
Clients
Ver 1.0 Hughes Proprietary Page 3 of 3
1.2. Managed Network Services: An Overview
The rate of technology change and the acceptance of technology as a
business enabler have served to complicate the role of enterprise IS
organizations. IS organizations are often tasked with managing
multiple, complex projects without the luxury of time to re-skill or
recruit employees for that challenge. New complex enterprise wide
applications and networking technologies offer the promise of more
agile, more competitive environments for the business, but installing,
maintaining and managing them is a daunting task. Moreover, CIO’s
are increasingly willing to offload day-to-day operations to external
Managed Network Service (MNS) Providers to focus on those activities
that are recognized as adding value to the business.
The results of a recent survey above which was conducted in the US
identify the core issues CIO’s are facing today with implementing
Network Management projects in-house.
Every customer wants to reduce the Total Cost of Ownership (TCO) of
his network. MNS allows the customer to achieve this goal by
providing:
• A Service Model, which takes care of High Capital Expenses
• Reduced Cost of Staffing, including training and retention Costs
• Shorter Project Cycle time
• Tight Control through Web Based Online Reporting
• Data Security
• Customized Services
Matching the right MNS Provider and the right type of engagement
model is important if enterprises are to get the most business value at
the “ right” price. The key differentiators between MNS Providers are in
their commitment to the business and the quality of relationship they
establish with their customers. These differentiators can be classified
under five major elements:
• Service line viability
• Technical Quality
• Ease of doing business
• Customer Satisfaction
Ver 1.0 Hughes Proprietary Page 4 of 4
• Service Level Guarantees
HECL Managed Network Services
HECL has been providing VSAT based WAN and networking services to
customers in India from 1993. With the advent of the Internet,
enterprise networks have become more complex, catering to increased
user demand for Quality of Service, both in terms of application
availability and response times. There are more connectivity choices
today, from Frame Relay and IP VPNs to leased lines and ISDN at a
reduced price point.
Service Offering
Clients
dB
CUSTOMER Server 1
WAN Mail
Server
Web
Server
dB Server 2
Management Data
Mercury
Web
Browser
Service Information Portal
Management Console
Event Browser
Reporter
Ver 1.0 Hughes Proprietary Page 5 of 5
1.3. MNS Center Technical Architecture
In such a complex environment, Enterprise Network Management
(ENM) has emerged as a key customer requirement. In line with its
philosophy of providing world-class services to its customers. HECL
has invested heavily into the Enterprise Network Management space
and set up a state-of-the-art MNS Center at Gurgaon, India. This 24X7
center monitors and manages customer networks pro-actively and
provides trend reports and recommendations at regular intervals. The
HECL MNS Center is connected to the customer central site via SCPC/
Terrestrial Link. This architecture provides the customer:
• Dedicated access to high level technical resources on a 24X7 basis
to monitor and resolve technical problems
• Sophisticated reporting & trend analysis and state-of-the-art QoS
and Application Performance Monitoring tools
• Managed Security Services
Ver 1.0 Hughes Proprietary Page 6 of 6
1.4. Infrastructure Details
HECL has been extremely careful in selecting the right set of hardware
and software tools at the MNS Center. A list of the infrastructure
investments is attached.
Infrastructure Details
Network Management Server: HP 9000 B Class RISC UNIX running
HP Openview
Service Information Portal: HP 9000 B Class RISC UNIX running
HP Service Information Portal
Reporting Server: HP Netserver 3000 LHr running Openview
Reporter
Internet Services: HP Netserver 3000 LHr running Openview
Internet Services
Data-warehouse: Embedded
CRM System: Talisma
Voice Switch: Nortel Meridien
LAN Switches : Nortel
Routers : Cisco
Firewall: Checkpoint
IDS: ISS RealSecure
Structured Cabling : Lucent
20
HECL’s approach to Network Management is to provide a true view of
the services relating to customers. This is achieved by defining the
relationships between the infrastructure components, so that the
service impact of any infrastructure event can be easily determined.
Also, the root cause of any service violation can be pinpointed.
The HECL MNS Center architecture is made up from a number of HP
OpenView , HNS proprietary and Cisco components:
• HP OpenView Network Node Manager provides in-depth views of the
network through an intuitive graphical format. Combining trend
analysis, threshold, and Java™-based web reports, with data
warehousing enables proactive network management.
Ver 1.0 Hughes Proprietary Page 7 of 7
• Openview Internet Services (OV-IS) server – the measurement
server where the service probes are configured, and the transaction
responses are collated for reporting. VP-IS performs both active
and passive monitoring.
• Openview Reporter – facilitates reporting of data providing a
customizable WEB reporting interface that have been integrated
with the existing reporting mechanisms
• Service Information Portal – Gives the customer a secure
personalized view into their managed environment
• HNS Proprietary Software - Provides the core management platform
for the HNS VSATs and DirecWay Broadband Services
• Cisco Works 2000 for managing Cisco devices on customer
networks
1.5. Network Management
The MNS Center provides in-depth views of the network through an
intuitive graphical format using HP Openview Network Node
Manager and device specific applications like Cisco Works 2000. It
discovers network devices and provides a map to illustrate what the
network actually looks like. The multi-level map indicates which
devices and network segments are healthy and which areas need
attention. When a flood of events from the failure of a major device
appears in the alarm browser, a powerful event correlation engine
evaluates the event stream and pinpoints the root cause of the failure.
Combining trend analysis, thresholding, and Java™-based web reports,
with data warehousing enables proactive network management.
The MNS Center immediately pinpoints the source of network problems
and assists customers in planning for network growth and designing
network changes.
Remote user access facility provides the HECL MNS staff flexibility to
access network management tools from anywhere on the web. Using
these powerful features, network administrators are able to more
intelligently manage the network, leading to increased network uptime
and lower costs.
a) NMS Architecture
The NMS system consists of a set of run-time applications based on
the simple network management protocol (SNMP). Additional
applications are installed on top of the base system, presenting
Ver 1.0 Hughes Proprietary Page 8 of 8
integrated management applications sharing a common graphical user
interface. The HECL MNS Center offers a comprehensive and growing
repertoire of tools to meet diverse integrated network and system
management needs.
The NMS framework provides an open, standards-based foundation on
which a wide array of specialized management applications reside.
Interoperability and consistency of operation characterize applications
developed within this framework.
b) Managing Network Complexity
The NMS has been implemented keeping a distributed and scaleable
architecture in mind which disperses UNIX or Windows NT collection
stations for user-specified enterprise segments called domains.
Collection stations report device discovery and change to one or more
management stations, even across wide area networks. This
distributed management model means that the HECL MNS Center
provides the flexibility to tailor a application to the organization's
specific network management needs.
The distributed discovery and monitoring approach lets HECL locate
processing closer to the local domains of the customer environment.
The enterprise IT environment can be completely managed through
collection stations and management stations. A designated collection
station can manage specific devices.
HECL MNS optimizes network use by reducing the volume of data
flowing to central management stations for processing. Collection
stations can be configured to initiate independent responses to specific
events, eliminating the need for management station intervention.
Only critical event data is passed on, allowing available network
capacity, processing power, and IT staff resources to be channeled to
the customer’s business information and operational needs.
c) Automated Discovery and Mapping
HECL MNS discovers TCP/IP, IPX (supported on NNM for Windows NT)
and level 2 devices on the network and presents this information in an
intuitive graphical format. Level 2 discovery includes devices that
support Bridge, Repeater/802.3, or MAU MIBs. It discovers Cisco
devices, and provides intuitive views of these switches, as well as
physical connectivity, attached device, port, and VLAN information.
Ver 1.0 Hughes Proprietary Page 9 of 9
The NMS automatically maps the UNIX and Windows NT network
segments and nodes in the customer’s LAN/WAN environment
delivering accurate and up-to-date views of the network. It
graphically displays the network topology as it actually exists,
including the current status of the intelligent peripheral (IP)-
addressable devices on the network. All control of the networked
systems is from a consistent view. The NMS continuously monitors the
network for new devices and for the status of equipment on the
network. In addition, HECL can easily integrate new and existing
applications into the NMS user interface and enterprise map.
d) Java-Based User Interface
An integrated Web Launcher allows HECL MNS Center staff to start
Java-based applications such as the Network Presenter, the Alarms
Browser, and the SNMP Data Presenter from anywhere. A log-in facility
with password authentication provides security for the management
data. The launcher also allows user roles to be defined and information
to be filtered based on a user’s management objective.
The Network Presenter and the Alarms Browser dynamically update
maps and events automatically—no intervention is required. Offering
the highest level of flexibility, network topology is presented in either
graphical or tabular format. In addition, the SNMP Data Presenter
allows users to query the network for SNMP data such as interface
traffic, CPU load, or traffic routing to assist with network diagnosis and
planning.
The NMS provides easy access to network maps and enables
management of data from anywhere via the Web. Up to 20 operators
monitor customer networks by distributing the user interface to
multiple workstations. These workstations are located at the HECL
MNS Center. The Windows graphical user interface provides
sophisticated capabilities for quick and easy task execution within the
system and the network environment.
e) Reducing Network Downtime
In order to keep the customer network operating, HECL's MNS staff
must react quickly when problems occur. Network Node Manager
offers the ability to resolve problems by allowing operators to
configure automatic actions based on thresholds. When an event
threshold is exceeded, operators are immediately notified via a pop-up
Ver 1.0 Hughes Proprietary Page 10 of 10
window display. Remote notification can also be initiated using a
modem or paging service.
Event categories may be prioritized to gather event information from
anywhere in the enterprise. Robust filtering allows HECL staff to focus
attention on events critical to resource availability and performance.
f) Graphical User Interface
Based on OSF/Motif, the Windows interface allows HECL staff to easily
monitor and control the entire network environment. The NMS allows
the following tasks to be performed effortlessly and without
programming:
• determine the operational status of network
devices by group or status category
• define map symbols and toolbar icons to
automatically initiate a program or shell
script
• define map symbols and toolbar icons to
automatically explode into multiple icons
representing complex objects
• integrate applications quickly while
providing a single point of monitoring and
control
• use the pan and zoom feature to focus on
critical areas of large submaps while
maintaining a high level view
• drag and drop objects for easy cut and
move operations
• initiate frequently used tasks with the
toolbar
• customize map menus and symbol pop-up
menus
• use Web interface for easy access to
network status and events
• access specific submaps quickly
• use the Hypertext Help System for fast
assistance
g) Fault-Tolerant Capabilities
Ver 1.0 Hughes Proprietary Page 11 of 11
HECL can schedule back-ups for business-critical network management
information while continuing to manage critical network elements.
Additionally, collection stations can be configured to “failover” to
management stations so that in the event of a collection station
failure, network monitoring will continue uninterrupted.
h) Event Subsystem
The Event Browser allows HECL's operators to quickly prevent or
analyze and resolve network faults. Recognition and resolution of
undesired events or conditions makes sure that HECL's network is
always at its optimal operating state. The event subsystem allows
operators to:
• Filter events by any combination of source,
message string, time and date received,
severity, or acknowledged status
• Employ color-coded event notification icons
and acknowledgments to make sure that
the most important events are identified
and handled first
• Customize event notification and alarms by
defining event categories and creating
additional operator actions
• Configure events on a per-node basis
i) Event Correlation Technology
HECL MNS identifies problems faster through an advanced event
correlation service (ECS) engine. The engine intelligently correlates
events into high-level alarms, immediately pinpointing the root cause
of network problems. The drill-down capability allows network
administrators to see the contributing events for each of the alarms.
Out-of-the-box correlation logic is already implemented for some of
the most common network management problems. The pre-configured
correlation logic leads to more efficient network polling based on the
latest topology data. A separate event correlation designer allows
development and testing of correlation logic for very specific
correlation requirements through an easy-to-use GUI. These logics can
then be deployed to any number of collection stations or management
stations in the environment.
Ver 1.0 Hughes Proprietary Page 12 of 12
j) Network Discovery
The discovery system of HECL MNS:
• Automatically generates and maintains a
map of the TCP/IP and IPX networks,
including level 2 devices
• Continuously discovers and monitors the
status of network nodes, even across wide
area networks and across noisy or busy
connections
k) Dynamic Data Collector and User-Defined Thresholds
HECL MNS collects network data and allows the definition of thresholds
that trigger events. This helps in the planning and management of the
network, ensuring fast and reliable performance. The data collector
allows HECL's operators to:
• Obtain real-time and historical network
information for efficient troubleshooting and
planning
• Manage thresholds on critical resources and
automatically receive alarm notifications
• Define collections and thresholds for
multiple devices simultaneously
l) Distribution and Scalability
HECL MNS helps efficiently manage the complexities of the customer
network environment whether it consists of small workgroups or large
enterprise domains. HECL can:
• use the management console feature to
distribute tasks among operators and
lighten the processing load on the
management station enabling the
management of larger networks. Up to 20
distributed operators can have a
personalized and unique view of their
respective management domain.
• off-load CPU-intensive applications from the
management station.
Ver 1.0 Hughes Proprietary Page 13 of 13
• zoom in and out on sections of large
submaps with the pan and zoom feature.
Windows NT and UNIX collection stations can be distributed
throughout the environment so that data can be collected locally and
then forwarded to one or more Windows NT/UNIX management
stations. Windows NT remote consoles connected to UNIX
management and collection stations allow multiple operator access at
a lower cost. The new Java-based Web UI increases the number of
supported operators and offers connectivity over WAN links.
m) Fault Management
HECL MNS offers the following fault management capabilities:
• find network routes using a graphical and
textual display
• diagnose network problems—test problem
nodes selected from the map or entered
manually
• document network topology—compare the
current network map against future
planned topologies
• perform real-time comparisons to
benchmark data
• customize the fault management
application—add programs or shell scripts
Configuration Management
• Locate Network Objects—locate objects via
attributes to improve inventory control.
• List Remote Network Services—provide
information about TCP/IP and IPX services
available on remote nodes.
• Edit Network Maps—non-TCP/IP/IPX nodes
can be manually added to the map.
• Configure Nodes—via the NMS
• Obtain Information—node-specific
information such as location, owner, and
configuration may be easily collected.
• Retrieve Management Information Base
(MIB) Information—point-and-click through
Ver 1.0 Hughes Proprietary Page 14 of 14
menus to obtain information on any SNMP
device.
• Keep Electronic Records—changes are
automatically detected and recorded.
Performance Management
• Monitor Network—collect statistics and
display them in a variety of formats.
• Monitor System Resources—monitor the
relative load of systems running an SNMP
agent.
• Determine Usage Profiles—using the data
collector, determine utilization profiles of
resources.
n) MIB Application Builder
The MIB Application Builder enables HECL to quickly build MIB
applications for standard and enterprise-specific MIB objects without
programming.
The MIB Application Builder provides a quick, non-programmatic way
to customize the NMS application allowing HECL to manage
information for specific needs. All of your IP-addressable devices can
be managed by loading vendor-specific MIB extensions through an
easy-to-use menu. Specific MIB variables can be easily controlled by
using the SNMP set command.
o) Customer Views
Customer Views extends the power of the NMS to provide intelligent,
customer-based management of network environments for HECL staff.
Customer Views enables HECL to associate network resources, such as
internet access links and servers, as well as network events, with the
customers that use and are affected by them. Network managers can
respond proactively, correcting the fault as well as contacting the
customers. This allows improved network topology design and
increased customer satisfaction.
Ver 1.0 Hughes Proprietary Page 15 of 15
Customer Views also enables HECL to make other logical groupings of
network resources, such as by location. It also expands the link
management capabilities of the NMS, including management of
individual interfaces and expanded information about Frame Relay,
ATM, SONET and DS-1/3.
1.6. HECL Service Management Strategy
The need for rapid response support is growing in direct proportion to
the phenomenal rate of change. HECL offers an unparalleled
combination of integrated support processes covering the operation,
performance and business intelligence associated with the seamless
management of the customer’s network infrastructure. HECL’s service
level management strategy integrates processes such as call
management, incident management and problem management,
together with quality management processes: configuration, change,
workflow and service level management. This approach ensures that
customer’s critical business services are well executed today and in the
future.
HECL’s service management strategy is a preventative service rather
than a reactive one. The exact impact of incidents is immediately
known and HECL is equipped with the right information and tools to
restore services before the end-customer experiences any difficulties.
A preventative service is critical to support MNS services because
every disruption business processes is immediately visible to the
customer. These disruptions can result in tremendous loss of revenue
for the customer.
HECL offers comprehensive service management capabilities
encompassing helpdesk, incident, problem and change management
processes into a single, seamless workflow. Through this unique level
of integration, the entire workflow process is implemented from the
work management process, which enables the MNS Center to operate
preventatively. With critical information at their fingertips, operators
and technical staff at the MNS Center can respond to and resolve
problems before they affect vital business services. HECL has
consolidated all critical service components into one service
management solution at its MNS Center. The solution is based on
Talisma CRM .The integrated services include:
• configuration management
• call management
• incident management
Ver 1.0 Hughes Proprietary Page 16 of 16
• problem management
• change management
• work management
• service level management
• Service Information Portal
This comprehensive approach to support helps HECL customers
maintain competitive advantage through optimized network uptime,
application availability and proactive support for their internal and
external customers.
p) Configuration Management
The MNS Center tracks and controls configuration items (e.g. hardware
inventory, IP addressing scheme, router configurations, IOS versions
etc) throughout their lifecycle. In addition to providing information to
other processes such as problem and change management,
configuration management also includes easy access to information
such as service contracts, relationships/dependencies between
configuration items and organization-related information.
q) Call Management
The heart of HECL’s Service Level Management strategy is managing
customer requests and incidents. Because optimizing the end-user
experience is a primary goal, HECL has chosen the Talisma Customer
Relationship Management solution, which offers a variety of
capabilities that streamline interactions with end users.
r) Managing incoming calls
HECL’s CRM set-up allows first and second line support engineers to
quickly resolve calls or assign them to a second line support specialist.
Through its integration, the CRM solution gives the first line support
specialist immediate access to other information such as known
incidents, problems or changes associated with specific infrastructure
components. Access to this information increases the number of calls
resolved on first contact, improving the productivity of HECL staff.
s) Incident Management
Incidents are a specific type of service call, usually referred to as
disruptions of service. Because of the seriousness of service
disruptions, HECL’s CRM set-up is bi-directionally integrated with its
Network Management system so that events are quickly and
Ver 1.0 Hughes Proprietary Page 17 of 17
accurately reported. Routing incidents to the CRM ensures that the
incidents are responded to in the appropriate order of priority.
t) Problem Management
Problem Management is often referred to as quality management
because the process focuses on analyzing calls and incidents to detect
a recurring pattern. These patterns identify structural problems within
the infrastructure and schedule them to be fixed. The internal
database of the CRM application as well the integration with external
knowledge bases helps HECL technical staff by identifying root cause
analysis.
u) Change Management
Change management has become increasingly important as the rate of
adoption of new technology continues to increase. HECL’s change
management process links the processes of initiating, scheduling,
assessing, implementing and evaluating changes to the network
infrastructure. The change management process is focused less on the
tools used to make the actual change and more on the management of
the change and its impact to the production environment. It is virtually
impossible to successfully manage a complex network infrastructure
unless operators have up-to-date information about what equipment is
actually in place at any given time.
v) Work Management
Service calls, incidents, problems and changes often result in a vast
amount of work to be done. Work orders are a tool for HECL staff to
plan, assign and follow-up on that resulting work. HECL has integrated
complete work order management and tracking at the MNS Center to
ensure work is completed as quickly and smoothly as possible
Details regarding the planned cost and maximum time expected for
task completion can be noted in the work order. The person initiating a
work order can specify a deadline for completion and assign a time
limit on the amount of time to be spent on the work. As the work
progresses, the work order can be updated with information reflecting
the actual completion time and date, any costs incurred or other
changes in the scope of the work. Views of the status of each work
order is available and adjustments to the scheduled work are allowed
as necessary.
Ver 1.0 Hughes Proprietary Page 18 of 18
An audit trail of completed work or work still in progress is provided in
a variety of views. Responsibility for completing a work order can
change. If the individual assigned the work order cannot complete the
task, it can be assigned to another specialist or reassigned to the
initiator of the work order. History lines are automatically created
whenever a work order status is changed. The history lines record who
changed the work order and when the work order was changed. By
viewing the history lines one can determine when a work order is
completed, overdue or stalled.
w) Service Level Management
By providing service levels based upon the requirements of the
customer’s business needs, HECL enables the customer’s IT staff to
make a significant contribution to their overall business.
Key customer benefits of service level management include:
• better balance of network benefits against network costs
• improvement in business performance and in customer perception
of the delivered managed network services
• reduction in the incidence of poor service for improved business
performance
• improvements in planning, scheduling and budgeting through the
introduction of processes
• decrease in the cost of missed business opportunities
• reduction in time needed to restore services
HECL MNS makes it possible to register and maintain services and
SLAs as well as multiple sets of support hours. Available support hours
are linked to the service levels used when calculating deadlines. The
priority level assigned to a service call is based on the related service
level agreement and the impact. The deadline calculations are based
on:
• the service level agreement in place for the customer
• the impact and priority settings for the service level
• the priority and the maximum duration settings
1.7. Service Information Portal
Service Information Portal contains the portal foundation for the
network and a range of management Information modules. The
Ver 1.0 Hughes Proprietary Page 19 of 19
modules extract information from the various management solutions
to provide an integrated view of ABIH’s managed environment. The
foundation provides administrative and server functions, as well as
three general-purpose modules. The three general-purpose modules
provide a flexible method for including data from a wide range of other
sources. SIP includes network management, application management,
operations management and reporting modules. HECL can provide
personalized views for individual AVB Group Company or for groups as
defined by Customer. Each view can contain one or more tabs for
grouping the information. Each tab then contains one or more modules
of information, as best suits the need of each company. SIP enables
Customer to easily filter the data from their managed environment so
that the end customer sees only information relevant to them.
Detailed SIP information includes network information such as
topology maps, events, performance summaries and network device
health, augmented with messages and other custom information. SIP
complements the network-operator-focused NNM web user interface,
providing value as a customer service tool with data meaningful to the
end customer. It consolidates information from all NNM subsystems
into one view, periodically extracting this data to update the customer
displays. It also filters this information so that each customer sees
information about their managed network environment, and not that
of other customers.
The SIP foundation provides several core functions that enable SIP to
extract and format information. The foundation provides:
· An Administrative GUI for creating and editing user views
· Managing security and login requests from users
· Requesting data from the individual modules
· Formatting that data and presenting it over the web
· Three general-purpose modules (Bookmarks, Generic, Message
Board), provide the basis for customer XML files.
Ver 1.0 Hughes Proprietary Page 20 of 20
1.8. Application Management
The Application Management components of HECL MNS provide a
mechanism to perform active tests of the services provided by the
HECL'S customers infrastructure, recording measurements for
response time, setup time, throughput and availability of the
application, and Policy Based Routing for Bandwidth management on
the links.
The basic architecture is built around the concept of a central
measurement server at HECL MNS, with distributed probe nodes at
customer locations. The configuration is defined centrally via a GUI,
and then distributed to the probe nodes. The service tests are
performed on the desired applications and target systems by the probe
systems, and then the results are forwarded to the measurement
server (via HTTP or HTTPS).
Service Level Objectives (SLOs) can be defined for various
measurements for each of the services, along with alarm definitions
for the generation of jeopardy alarms (alerts prior to the actual SLO
being violated).
The alarms are passed into the event management system via a local
agent and the response times and service level performances are input
into the performance sub-agent for ongoing storage and reporting.
Ver 1.0 Hughes Proprietary Page 21 of 21
Reports showing current and historical performance are then
presented via a Web interface on the measurement server.
1.9. System and Operations Management
HECL MNS solution based on the OpenView platform is a distributed
client/server management solution that provides operational and
performance monitoring for the enterprise. The architecture consists of
a central configuration and monitoring server, with distributed
intelligent agents.
The configuration of the agent is performed at the central server, with
templates being defined via the GUI for the various message sources
to be monitored. The templates are then assigned to the relevant
servers, and the system automatically deploys the monitors and
commences monitoring of the requested sources. Matching event
conditions on the agent are then escalated back to the management
server for viewing and action by the relevant operations staff.
It should be noted that the agent is not reliant on the management
server to filter non-important messages. The distribution of the
templates enables the agent to make local decisions as to what is
important, and escalate only relevant messages to the server.
Automated actions are available such that the agent can take remedial
actions without requiring interactions with the server.
The performance sub-agent monitors the health of the operating
system, analyzing and recording over 300 metrics. The performance
data is stored in a compact binary data-store on the client. Metric
thresholds can be defined such that performance alarms are generated
via the Operations sub-agent. Performance data can be viewed using
the PerfView viewer on the server, or data can be exported into
standard report files for additional analysis. The capabilities of the
performance sub-agent are discussed in more detail later.
The key capabilities of the event management system are
characterized in the following topics:
x) Message Groups
The use of message groups provides the mechanism to classify alarms
into different categories. This provides the operator with a clear view
of where the alarm is occurring and what the significance is. Examples
Ver 1.0 Hughes Proprietary Page 22 of 22
of the message groups provided are: OS, Performance, Oracle
performance, etc.
The message groups are fully customizable, and also provide the
capability to segment, which operators see, which alarms.
y) Applications
The OV configuration provides a mechanism to configure application
interfaces for operators. This mechanism allows system administrators
to provide only specific access to hosts for operators, and enables the
access without having to allow full system access. The ability to define
and assign applications to operators has a number of advantages:
firstly providing the operator with a level of abstraction from having to
know exactly how to execute commands; and secondly, providing
additional security within the environment.
z) Templates
The template configuration provides administrators with the ability to
define the monitoring parameters in a central location, and then
distribute these configurations out to the intelligent agents for
execution. The templates are defined via the GUI, with no coding
required. The templates are then assigned to the relevant nodes either
directly (1-1 template – node mapping) or via a node group
(functionality based – all WEB servers get the WEB templates). This
assignment significantly simplifies the configuration requirements to
enable monitoring on the distributed agents.
The ability to define and then assign templates based on functional
role provides HECL MNS with a rapid deployment mechanism.
Standard templates can be defined for various functional roles (WEB
server, DB server, etc), and then when new servers are deployed, the
standard templates can be pushed to the agent for instant application
monitoring.
Furthermore, operator instructions and actions can be defined as part
of the templates, enabling the operations staff the use the event
management interface to gain instructions on how to deal with
particular events. This significantly reduces the training requirements
for operators, as the problem resolution processes can be integrated
into the event management environment.
Ver 1.0 Hughes Proprietary Page 23 of 23
aa) Operator Roles
The event streams can be segmented such that operators only see the
events relevant to their functional role. This assignment is done based
on Message group and node group assignments, and ensures that
operational activities can be restricted to those operators authorised to
view/resolve events.
bb) Navigator View
The Navigator view (example below) provides the operations team to
define a full service hierarchy, with events and actions mapping into
particular points within the service tree. The service hierarchy provides
a view of the infrastructure being managed, along with relationships to
the services being provided by that infrastructure and ultimately the
customers using that infrastructure.
The benefits of using the navigator view is that the operator is
immediately provided with an impact assessment of a particular event,
being able to see which services and customers are impacted by an
outage. Root cause analysis is also available, enabling an operator to
easily view what events are currently impacting a customer or higher-
level service.
The service hierarchy is fully customisable, enabling HECL MNS to
define a model that accurately maps the services being provided.
cc) Event Correlation
Event correlation mechanisms are provided for the correlation of
system and network events at a number of levels within the solution.
The Event Correlation Services (ECS) are provided for integration in:
1. NNM (correlation of network specific events)
2. OV agent (correlation of events within the agent),
and
3. OV server (correlation of events within the
server).
Ver 1.0 Hughes Proprietary Page 24 of 24
As discussed previously, the information above is a subset of the
functionality – please refer to the concepts manuals for more detailed
information.
dd) Scalable Architecture and Multiple Management Servers
The customer environment can be configured hierarchically and then
management responsibility spread across multiple management levels
according to criteria as diverse as operator expertise, geographical
location, and the time of day. This allows operators to focus on doing
what they are good at, safe in the knowledge that they have round-
the-clock technical support available either automatically or on
demand. HECL MNS flexible management capability using the
Openview platform provides customer with the means to:
• Manage your worldwide network more effectively, for example, by
using the follow-the-sun functionality .
• Increase efficiency by implementing competence-center policies.
Escalate messages to different servers, and forward messages
between management servers.
• Manage an expanding network environment and reduce primary
server overload. The potential bottleneck that results from having
all managed nodes report to a single management server (and the
consequent database-related performance issues) is eliminated if
managed nodes can report to different management servers.
ee) Back-Up Management
HECL MNS back-up management solution is designed around the HP
OpenView Omniback II platform. The solution is capable of handling
the pace of the customer's storage growth. Whether the backup
requirements are for a single system environment, a small workgroup,
or a multi-site heterogeneous data center environment with thousands
of servers, Omniback software provides the centralized backup and
speedy recovery for Customer's business demands. At the same time,
Omniback delivers the comprehensive scheduling and automation
Customer needs to protect more data within the same people-print.
Ver 1.0 Hughes Proprietary Page 25 of 25
HP OpenView Omniback II utilizes a broad range of techniques to keep
data available, even during backup. From online and open-file
backups, to online incremental backups (Oracle ® ), to clustered
backups, server-less backups, all the way to zero-downtime, zero-
impact backups, Omniback offers a solution to meet Customer's
application needs. Using innovative split-mirror techniques, Omniback
enables fully integrated solutions for automated, non-disruptive, point-
in-time backup covering the broadest set of storage
configurations,operating environments and applications. So while
Omniback is protecting Customer's most valuable asset—its data—it is
also keeping the data available so that applications remain running
and ready for business.
Of course, backup is important, but it isn’t until one has lost
information and needs to restore that one realizes the importance of
having a reliable data protection and recovery solution. Omniback
provides an equally comprehensive and flexible recovery solution.
Whether browsing a directory listing to restore a single file, an
Exchange mailbox, a database table, or in full bare-metal disaster
recovery mode, restoring data is fast, simple and efficient. Simply put,
Omniback software offers the industry’s most reliable, scalable, high-
performance data protection available for heterogeneous computing
environments.
The Omniback solution provides a single, consistent interface to
automate and monitor backup and recovery processes. Omniback
supports a comprehensive list of backup clients (disk agents) and
backup device servers (media agents) to extract application data and
protect it onto backup drives and tape libraries. Additional database
and application agents provide support for specialized backup
approaches such as online, open-file, serverless or zero-downtime
backup integrations, allowing to select the particular protection
approach for Customer's application needs.
For optimum flexibility with maximum control, backup management
can be distributed or centralized. At each location, a management
station (cell manager) schedules, initiates, and monitors the backup
sessions. Information about files, hosts, backup sessions and media
are carefully tracked in a central database to enable swift identification
of file versions available for restore, and which media are located
online or in an off-site vault. For multi-site operations, a manager-of-
managers capability distributes control to local administrators while
maintaining the ability to set overall policies and monitor the entire
Ver 1.0 Hughes Proprietary Page 26 of 26
enterprise backup environment. As a result, corporate policies can be
implemented without diminishing local control and responsibility.
Automated library and improved media management, event
notification, advanced scheduling capabilities and the ability to
automatically manage remote and local backups through a single,
central management station means you’re able to offload routine
backup tasks and apply scarce IT expertise to critical strategic
projects. Even installation can be automated—Omniback provides
built-in software distribution capabilities enabling remote,
simultaneous installation and maintenance of backup clients and agent
soft-ware across the network. With Omniback, flexible backups with
one single GUI across platforms is possible. This helps to simplify
operational tasks and cut down on training time. Omniback also
provides comprehensive insight into the backup system and process.
Automated health checks monitor the integrity of the backup
environment to ensure reliable backup completion.
Omniback ensures that the customer is notified of significant events
via event-based and scheduled notification through e-mail, SNMP
traps, event log, network broadcast or pager. As a result, the
customer has the information he needs when, where and how he
needs it. If one needs to access the system, sophisticated Web
reporting allows easy viewing of the status of backup jobs from any
Web-accessible location.
data -print.
•
Ver 1.0 Hughes Proprietary Page 27 of 27