1、 TECHNICAL REPORT ATIS-0100004 AVAILABILITY and 2) Restorability. The purpose of the ETS is to facilitate emergency recovery operations for restoring the community infrastructure and for returning the population to normal living conditions after serious disasters and events, such as floods, earthqua
2、kes, hurricanes, and terrorist attacks. The ETS will be provided through shared resources from the public telecommunications infrastructure, which is evolving from a basic circuit-switched configuration of todays conventional telephone networks to an Internet-based, packet-switched technology provid
3、ing a richness of communication capabilities. The timely establishment of an effective ETS has been given significant urgency as a result of the September 11, 2001 terrorist attacks in the United States (U.S.). Many challenges and considerations need to be addressed in defining and establishing the
4、functional capabilities for the ETS in the emerging packet-based telecommunications services. This TR addresses the requirements: d. Restorability and n. Reliability/Availability of the list of the eighteen (18) requirements in T1.TR.79-2003 1 for ETS that packet-based telecommunication and wireless
5、 mobile networks are capable of providing, and that must receive attention during the process of the convergence of these technologies. Disaster situations can occur unexpectedly at any time, any place. These events often significantly damage the community infrastructure and severely disrupt daily l
6、iving. Recovery requires rapid response by local authorities, immediate reaction from utility service providers, and support from medical, construction, fire, and police resources. Effective communications are essential to facilitate the myriad activities for coordinating lifesaving activities ATIS-
7、0100004 2 concurrent with reestablishing control in the disaster area. Following a disaster, immediate response operations focus on saving lives, protecting property, and meeting basic human needs. 1.3 Application As stated above, this TR is meant to be applied by those standards bodies, particularl
8、y ATIS Technical Committees, in their work on current and future enhancements to evolving networks. It is hoped that it will provide much-needed guidance concerning where issues are being addressed relating to ETS requirements. It is not meant to imply that certain requirements will be mandated by l
9、aw. 2 NON-NORMATIVE REFERENCES At the time of publication, the editions indicated were valid. All standards and reports are subject to revision, and parties to agreements based on this Technical Report are encouraged to investigate the possibility of applying the most recent editions of the standard
10、s and reports indicated below. 1 T1.TR.79-2003, Overview of Standards in Support of Emergency Telecommunications Service (ETS).12 ITU-T Recommendation E.106, Description of an International Emergency Preference Scheme.23 reference deleted 4 Telcordia SR-2275, End-to-end availability objectives for l
11、ocal exchange networks on reference architectures and hypothetical reference networks.25 Jones, V., High Availability Networking with Cisco, Addison Wesley Longman; 1st edition (December 21, 2000). ISBN 0-201-70455-2 6 “Executive Overview,” SLA Management Handbook, Volume 1, TeleManagement Forum, Me
12、mber Evaluation Version 2.0, July 2004.37 “Concepts and Principles,” SLA Management Handbook, Volume 2, TeleManagement Forum, Version 2.0, April 2004.38 “Service and Technology Examples,” SLA Management Handbook, Volume 3, TeleManagement Forum, Member Evaluation Version 2.0.39 T1.TR.55-1998, A Techn
13、ical Report on Reliability and Survivability Aspects of the Interactions Between the Internet and the Public Telecommunications Network.110 T1.TR.70-2001, A Reliability/Availability Framework for IP-based Networks and Services.1 11 ATIS-0100006, Service Restoration Priority Levels for IP Networks.11
14、This document is available from the Alliance for Telecommunications Industry Solutions (ATIS), 1200 G Street N.W., Suite 500, Washington, DC 20005. 2Telcordia documents are available from Industry Direct Sales, Telcordia, 8 Corporate Place, PYA 3A-184, Piscataway, NJ, 08854-4156, or: . 3Available fr
15、om the Telemanagement Forum (TMForum) at ATIS-0100004 3 3 DEFINITIONS, ACRONYMS, instant messaging; wireless access; unicast/multicast/broadcast of audio, video, and data; interactive video; remote database access; etc.). 5 AVAILABILITY ASPECTS OF ETS ETS can be implemented in both Virtual Private N
16、etworks and Public Access Networks. These two types of networks can be used to support a wide range of ETS applications such as Voice over IP, Short Messaging, Instant Messaging, Videoconferencing, Wireless Services, Database Access, etc. IP Networks carrying ETS traffic should have redundant routes
17、 (e.g., multihoming, route diversifications, etc.) and be properly engineered to ensure ETS calls and/or sessions can go through during network congestion. ATIS-0100004 8 5.1 Availability Calculations Availability measures the percentage of time the service system is in its operational state. To ass
18、ure high availability of service, failures of each service system (i.e., MTBF) must be infrequent, service system down time (i.e., MDT) must be kept small, and service recovery (MTTR) must be expeditious (once there is a failure or outage). MDT = MTTR + MLDT A = MTBF/(MTBF + MDT) Where A is Availabi
19、lity, MDT is Mean Down Time, MTTR is Mean Time To Repair, MLDT is Mean Logistical Down Time, and MTBF is Mean Time Between Failures. As noted in 10, “network outages, such as network node or link failures may or may not have an impact on a network service.” If adequate provisioning is in place, a ca
20、tastrophic hardware failure might not result in any service outage. The following are examples of two end points (EPs) communicating through different networks and the associated availability equations. 5.1.1 Networks in Series Given the network A and network B in series, let the availability of net
21、work A be A1and that of network B be A2; the composite availability between two EPs is A1* A2. 5.1.2 Networks in Parallel In the network below, A is in series with B, network C is in series with D and the two series networks are in parallel with each other. The availability of A is A1, of B is A2, o
22、f C is A3, and of D is A4. The availability of the composite network transporting traffic between two EPs is given by (A1*A2+ A3*A4 A1*A2*A3*A4). ATIS-0100004 9 5.2 Availability Calculations in the SLA Handbook The SLA Handbook 789 provides guidance on the specification of service availability in Se
23、rvice Level Agreements (SLAs). The following equation from the SLA Handbook 9 defines service availability (SA) for leased lines. The assignment of different weighting to different network elements and to different measures of degradation (i.e. Severely Errored Seconds and Errored Seconds) allows fo
24、r flexibility in definition and measurement of Service Availability. SA is Service Availability, SDF is Service Degradation Factor, UAS is Unavailable Seconds, SES is Severely Errored Seconds, ES is Errored Seconds, Wkis the weight for Service Access Point (SAP) k. A SAP is defined in the following
25、figure from 9. Figure 2 - Network example showing Service Access Point (SAP) ATIS-0100004 10 “The data terminating equipment (DTE) and data circuit terminating equipment (DCE) are intended to be functional descriptions. The DCE and the DTE may be physically implemented in the same or in different ha
26、rdware devices.” 9 The SLA Handbook contains several other equations for calculating Service Availability, including some which take into account the users perception of a degradation of service. 5.3 Availability Objectives for ETS As part of a service-level agreement between a service provider and
27、its ETS users leasing a connection, the carrier must commit to providing certain availability for the connection. A common goal for end-to-end connection is around 99.9% (three 9s) of the time. This goal corresponds to a connection downtime of less than 9 hours per year as indicated in the table bel
28、ow. A connection is routed through many nodes in a network between its source and its destination, and there are many elements along its path that can fail. In order to obtain 99.9% of service availability for ETS, a service provider must properly design and engineer its network to ensure it continu
29、es to provide service (as specified in the SLA) in the presence of failures. Below is the availability range applicable to communications systems and associated downtime: Availability Downtime per Year Seconds Minutes Hours Days 99.9999% 32 99.999% 315 5 99.99% 3154 53 99.9% 31536 526 9 99% 315360 5
30、256 88 4 The use of redundancy (logical, physical, or both) is typically used to provide the necessary availability for ETS and related services. High availability can also be achieved through other means, such as the use of diverse routing and route weighting. 6 RESTORABILITY ASPECTS OF ETS Work on
31、 establishing restoration priority levels is under active study. (See ATIS-0100006.) ETS should be given the highest restorability class available. It is assumed that restoration schemes will be in compliance with FCC regulations (e.g., TSP). Within a single providers network restorability of ETS sh
32、ould be addressed in a Service Level Agreement (SLA) between the ETS service provider and the ETS service customer. Restorability applies to both service restorability and restorability of the physical layer. Examples of fast restoration mechanisms for possible use in ETS services are the following:
33、 Dedicated 1+1 protection scheme: With a 1 + 1 scheme, restoration can be done in 60 milliseconds. Example from IETF MPLS GMPLS: A shared mesh protection. Virtual Ring. Intradomain routing protocols: ATIS-0100004 11 o Intradomain routing protocols such as Open Shortest Path First (OSPF) and Intermed
34、iate System to Intermediate System (IS-IS). o Interdomain routing protocols such as Border Gateway Protocol (BGP) can facilitate resotoral across network domain boundaries. o Routing policies can be applied that influence how ETS packets can be routed in an IP network. ATIS Committee TMOC addresses
35、cooperative network management aspects involving multiple network operators which include restoration of network services. 7 EXAMPLE REFERENCE MODELS The following reference models are offered as examples. Combinations of these models will make up the environment for the ETS service. SLAs might be m
36、ade between different service providers and calculations of the individual availability will be used to determine if the overall availability will be sufficient to meet the requirements for the ETS. If ETS users expect 99.9 (for example) for end-to-end service across two networks, the two involved p
37、roviders have to design their networks and work out some network peering agreements with each other to ensure ETS users would get the availability for which they contracted. If the availability of one network is X and that of another network is Y, then the end-to-end availability is X*Y. For example
38、, given a network X availability of 99.94 4, in order to meet an end-to-end availability of 99.9 for ETS calls or sessions which traverse through a combined X and Y network, the availability of the Y network must be equal or greater than 99.4/99.9. It is possible that different network types would h
39、ave different availability requirements. For example, an ETS call traversing from an IP-based network to a wireless network might only be able to provide a composite availability less than 99.9 because of the non-ubiquity of the IP network and the potentially spotty coverage of the wireless network.
40、 7.1 Packet-Based to Packet-Based 7.1.1 Scenario 1: IP-based services provided by a single provider ATIS-0100004 12 7.1.2 Scenario 2: IP-based services provided by two providers 7.1.3 Scenario 3: IP-based services provided by three or more providers 7.2 Packet-Based to Wireless Network ATIS-0100004
41、13 7.3 Packet-Based to PSTN 7.4 Packet-based to 802.11 Wireless Network 7.5 PSTN to 802.11 Wireless ATIS-0100004 14 7.6 Packet-based to IP Cable Network 7.7 PSTN to IP Cable Network 7.8 IP Cable Network to Wireless Network ATIS-0100004 15 7.9 PSTN to Wireless Network 8 CONCLUDING REMARKS This TR ill
42、ustrates some internetworking scenarios where ETS can be offered to authorized emergency personnel (e.g., NS/EP community) during natural and/or man-made disasters. There are two different kinds of ETS applications: elastic and inelastic. The TR also provides some basic calculations for availability and suggests different service classes for restorability of ETS. Those network providers who offer ETS need to be able to deliver specific levels of network performance in support of these applications.
