RE: call numbers for tommorow 1pm est

["Lai, Wai S (Waisum), ALSVC" <wlai@att.com>]

To aid discussion, I have prepared an outline of some
potential topics that might be considered for inclusion
in the team's draft.  (Note that this is not intended
as an agenda.)
Thanks, Wai Sum.

-----Original Message-----
From: Ed Kern [mailto:ejk@tech.org]
Sent: Tuesday, June 19, 2001 2:26 PM
To: tewg-dt@ops.ietf.org
Cc: jimpb@sc.rr.com
Subject: call numbers for tommorow 1pm est


-----BEGIN PGP SIGNED MESSAGE-----



Confirmation # is 533-6213.    The phone code for the conference call is
6213.  The US number is the 1-888-467-2119 number.  What the form refers
to as "Local" is actually the internation number.  International would
dial
their countries code for the US _ _ _ -703-871-3713 followed by the conf
code 6213.  If any problems are encountered supposedly an operator is
available to help them. Just in case Genesys Conferencing number is
1-800-
777-1826.


Ed
-----BEGIN PGP SIGNATURE-----
Version: Mulberry PGP Plugin v2.0
Comment: processed by Mulberry PGP Plugin

iQCVAwUBOy+ZHOmO/gK7ZGvVAQEAnwP9Gaej6hEeDiuzcHeNPgEdhYevKf+xVBnY
jXkNPyLqNustImoZHPuXB5ahJopKpY1XalHqn4z58iADLNy15ApplB9qlXwaEx6z
nSsWRWpM0G8Zwv1hihgH20yvCIz6ml9cyUJEJFZprGurFihdXUhInVgb28NdyCb2
QYQDUQoMS/g=
=ja4t
-----END PGP SIGNATURE-----

Multilayer survivability.doc

Network Hierarchy and Multilayer Survivability
(A straw proposal)

Introduction
Objective: to ensure interoperability of protection/restoration mechanisms between multi-vendor equipment in a service provider network.
Focus: intra-domain operations

Definitions
Network hierarchy
Protection
Restoration
Network survivability refers to the capability of the network to maintain service continuity in the presence of faults within the network [TE-Framework].

Service requirements
Objective: (1) Enhance network integrity (i.e., reliability and maintainability) by using mechanisms and adopting policies to minimize the vulnerability of a network to service outages due to any hardware/software/procedural faults, failures or errors.  (2) Minimize impact to services and network performance within given constraints (e.g., minimize the impact due to the traffic imbalances created by network failures).

Minimize packet loss and packet mis-sequencing during the recovery period.

Best-effort data: restoration of network connectivity; rerouting at IP layer would be sufficient
Premium data service: (TCP or application protocol timer requirement?)
Voice: call cutoff is in the range of 140 msec to 2 sec
Other real-time service (streaming, fax?)
Mission-critical traffic

Differentiation of service reliability - the capability to specify and implement a different protection/restoration strategy (e.g., restoration priority, restoration time) for each class of traffic.

Operational requirements
Maintain a network's ability to operate at a stable state (e.g., automated control capabilities that will allow a network to adapt quickly to any significant changes in its state).
Support network maintenance operation by allowing service providers to follow their desired practices

Types of network failures
What types of faults exist?  What is the effect of a fault type on traffic?
Fault detection and coverage: different types of failures may be best handled by recovery actions at a certain layer.  Which layer/what mechanism is responsible for recovering from which fault?
Link failures: fiber cuts, transmitter failures
Node failures: processor or line card failures, memory failures, power failures, procedural errors such as misconfiguration

Layering Principles and Protection/restoration granularities
Time scale of operations: the closer to the fault a recovery action is performed, the faster the recovery
Recovery unit: virtual circuit or path, link, wavelength, fiber (the higher the layer, the finer the granularity)
Traffic unit: per flow, per diffserv class, per traffic trunk

Protection/restoration at different layers
Current network hierarchy: IP, MPLS/ATM, SDH/SONET, optical network
Protection and restoration schemes are being developed for each of these layers.
Functional distribution: physical layer impairments and fault isolation more easily handled by lower layers; protocol errors, node malfunction and rerouting more easily handled by higher layers.
Benefits: each layer does its best - enables a faster response to failures, thereby increasing network reliability
Features, advantages, disadvantages, issues, requirements for each layer.

Coordination among layers
Escalation strategy
Failures within a layer can be guarded against by techniques either in that layer or at a higher layer
Potential benefit: shielding of high layers from transport failures
Inter-layer communication requirements: especially when lower-layer restoration takes longer time than higher-layer restoration; use of hold-off timer to avoid contention between the different single-layer recovery schemes (how to choose timer values and what are the requirements)
Optimization of spare resources: improve resource utilization by sharing spare capacity across different layers: Sharing of spare capacity to minimize the amount of spare capacity required at each layer

Evolution towards IP over optical
Issues, requirements

Survivability Measurements
For each layer, define a set of measurements to characterize the survivability performance of a network.

Protocol requirements
Minimize the state overhead in maintaining at each node the recovery information (e.g., number of backup paths, level of coverage desired) to enable a switchover in the event of a fault
Routing protocol extensions
Signaling protocol extensions, e.g., fault notification, initiation of recovery operations (signaling speed requirement?)

Conclusions

References