Transcript Managing Enterprise Networks

Rmon
Diagram of RMON MIB
Root
ISO
Org
DoD
Internet
Mgmt
Private
MIB 1 & 2
RMON
1. Statistics
2. History
3. Alarm
4. Hosts
5. Host Top N
6. Matrix
MIB 1
7. Filter
8. Capture
MIB 2
9. Event
10. Token Ring
RMON MIB Groups
Statistics - Traffic and error rates on a segment
History - Above statistics with a time stamp
Alarm - User defined threshold alarms on any RMON variable
Hosts - Traffic and error rates for each host by MAC address
Host Top N - Sorts hosts by top traffic and/or error rates
Matrix - Conversation matrix between hosts
Filter - Definition of what packet types to capture and store
Packet Capture - Creates a capture buffer on the probe that
can be requested and decoded by the management application
Event - Generates log entries and/or SNMP traps
Token Ring - Token Ring extensions, most complex group
RMON Increases Management Capacity to 250%
Distributed Techniques Save Time and Money
Using RMON, Network Management Staff Can
Provide Services to More Users and Segments
RMON
segments
No RMON
200+
101 to 200
51 to 100
11 to 50
0
5
10
15
20
25
30
segments managed per staff person
Source: McConnell Consulting, Inc. 9/94 Survey of LAN Managers
Network Manager Needs More Help
Client / Server Revolution is Here
• Mission critical client/server applications mean that
network up-time and performance are required
• Good information currently available for segment traffic,
performance and utilization (RMON1)
• End-to-end, global view of enterprise traffic is needed for
troubleshooting today’s complex internetwork(RMON2)
– Isolate cause of problem quickly and respond
– Redeploy resources for optimal performance
– Spot bandwidth utilization trends before
bottlenecks occur
Remote Monitoring in the ISO Model
Going Up-the-stack With RMON2
Application
Presentation
Session
Transport
Network
Data Link
Physical
RMON2
RMON
Enterprise Network Traffic Management
Benefits of Monitoring Protocols “Up-the-stack”
• Understand network performance from
an application perspective
• Logical view of workgroup and application
communication patterns
• Clear visualization of end-to-end traffic
• High-value enterprise traffic information results:
– Improved performance through network tuning
and optimal placement of network resources
– Trend analysis based on actual growth rates
– Accurate accounting data
IETF Responds: RMON2 Standard
for Monitoring Higher Protocol Layers
• RMON (1) accepted and implemented
• RMON2 Working Group kicked off July 1994
• RMON2 Internet Draft delivered June 1995
• First RMON2 interoperability testing occurred
in September 1996
• RMON2 received its RFC #s (2021, 2074)
in January 1997
• Vendors can commence development of
RMON2-compliant implementations immediately
What is RMON2? Standard for
Monitoring Higher Protocol Layers
• Major new capability: Provide statistics
on network-and application-layer traffic
– Open structure for collecting traffic data
at higher protocol layers
– Protocol directory critical aspect of the MIB
• Additional enhancements:
– Address Translation
– User-defined histories
– Probe configuration-device, modem and trap
administration based on the Aspen MIB
– Time sorted tables
Diagram of the RMON2 MIB
Root
ISO
Mgmt
MIB 1&2
MIB 1
RMON
Org
DoD
Internet
Private
RMON1
1. Statistics
2. History
3. Alarm
11. Protocol Directory
12. Protocol Distribution
13. Address Map
4. Hosts
14. Network-Layer Host
5. Host Top N
6. Matrix
MIB 2
RMON2
7. Filter
8. Capture
9. Event
10. Token Ring
15. Network-Layer Matrix
16. Application-Layer Host
17. Application-Layer Matrix
18. User History
19. Probe Configuration
20. RMON Conformance
RMON2 MIB Groups
11
Protocol Directory - List of protocols the probe can monitor
12
Protocol Distribution - Traffic statistics for each protocol
13
Address Map - Maps network-layer to MAC-layer addresses
14
Network-Layer Host - Traffic statistics to and from each discovered host
15
Network-Layer Matrix - Traffic statistics on conversations between pairs of
discovered hosts
16
Application-Layer Host - Traffic statistics to and from each host by
protocol providing insight into the use and growth of applications
17
Application-Layer Matrix - Traffic statistics on conversations between pairs
of hosts by protocol
18
User History Collection - Periodic samples of user-specified variables
19
Probe Configuration - Remote configuration of probe parameters
20
RMON Conformance - Requirements for RMON2 MIB conformance
RMON2 Implementation Options Two Alternatives
Balancing Price and Performance
Network Layer
Application Layer
Statistics
X
X
Hosts
X
X
Matrix
X
X
Matrix topN
X
X
• “Type A” Implementations
– Network layer traffic
– Less memory and
processor power
– Expect to see embedded
in hubs and switches
• “Type B” Implementations
– Application layer traffic
– More memory and
processor power
– Basis for high-end traffic
management applications
Impact of RMON2: Adding a Logical View
to Management Products
• Answers critical questions like:
– What percent of WAN traffic is due to my
order-entry application?
– How fast is the Lotus Notes traffic growing?
– How many hours does Fred spend on the web?
• Requires new user-interface
– Scaleable, end-to-end view of all traffic on network
– Support range of logical groupings
– Linked to other management applications
• Essential for roll-out of client/server applications
Example RMON2 Solution: Transcend® Traffix™
Manager and SuperStack® II Enterprise Monitor
Tuning the Network to the Business
• Comprehensive understanding of network traffic
– Real-time and historical
– Applications and protocols
– Utilizing RMON2 data
• Optimize the network to meet the needs
of business applications
• Speed troubleshooting at the enterprise network
level, rather than the device or segment level
• Set real policies for the business use
of the network
System-level Management of Networked
Apps
• Top down view of all traffic on network
– End-to-end display of conversations
– Network-and application-layer traffic details
– Designed and tested to scale to large, complex networks
• Powerful, alternative logical groupings
– Much more than “Protocol Domains”
– Group by geography, function, subnet, VLAN, etc.
– Easy-to-use navigation
• Speeds troubleshooting
– Global view to see interaction and scope
– Zoom to problem hosts, devices, protocols
– Launch detailed protocol analysis on any link
Communicate to Users and Management
• Fully integrated traffic database
– Real-time and historical information in all views
– Histories over the past day, week, month,
year, etc.
– Trend analysis, comparison and graphing
– Set thresholds on key parameters
• Automated reporting
– Key trends in application, network
and device utilization
– Easy report set-up from same view as GUI
– Built-in, no data export/import required
Traffic Management is a New Discipline
Business ApplicationsTrafficInfrastructure
• Brings together:
– Application level monitoring
– Enterprise-wide instrumentation
– Ubiquitous standard
• To achieve:
– Analysis of applications’ use of infrastructure
– Performance management as seen by users
– Tuning, capacity planning, accounting, etc. based
on business applications usage
– System level troubleshooting
• In the face of:
– New network technologies (VLANs)
– Unpredictable traffic trends
Mission: Enterprise Traffic Management
Build on Foundation of Infrastructure Management
Business
Objectives
Quality of Service
Efficiency
Policies
Traffic Management
Survival!
Infrastructure Management
Troubleshooting
Tuning and Planning
Advantages: Enterprise Traffic Management
New Level of Network Knowledge and Capabilities
Activity
Infrastructure Management
Concepts and Physical - devices and
segments
Perspective
Trouble
shooting
Planning
Tuning
Policies
Device control and
configuration
Physical connectivity
See effects on devices
“Purchase Order” planning
Local utilization and rate of
growth at packet level
Standardize equipment
purchase
Traffic Management
Logical - traffic and applications
Reflects organization, scope
and groups
Behavior of whole network
Discriminate between application
problems and network problems
See causes, as well as effects
Business planning based on real
utilization and rate of growth
of applications
Define “reasonable” use
Set use policies and enforce them
Results: Enterprise Traffic Management
Improved Responses to Performance Problems
Add
Resources
Year
Reconfigure
Resources
Traffic
Policy
Trend/Report for Re-Design
and Adding Resources
Month
Days / Weeks
Hours / Minutes
Performance Assessment
for Tuning and Optimization
Prioritize Network Use
to Business Needs
Benefits: Enterprise Traffic Management
Match Network Investments to Business Objectives

$
– Improve network performance, eliminate bottlenecks,
so business applications run faster, more reliably
– Find enterprise network system problems fast
– Only invest due to business application growth
– Don’t waste expensive network resource
on “chat” traffic
– Efficient distributed techniques bring the problem
to the expert, rather than the expert to the problem
– Ease-of-use means information can be handled
by less skilled staff members