INTERNET-DRAFT R. Huang Intended Status: Standards Track Huawei Expires: December 30, 2013 June 28, 2013 Use Case for Large Scale Measurements Used in Data Collection of Network Management Systems draft-huang-lmap-data-collection-use-case-00 Abstract This document augments the use cases of large scale measurement of broadband performance (LMAP). It discusses measurements for a common platform which works for different usages including troubleshooting, performance understanding, quality evaluation and network adjusting. Status of this Memo This Internet-Draft is submitted to IETF in full conformance with the provisions of BCP 78 and BCP 79. Internet-Drafts are working documents of the Internet Engineering Task Force (IETF), its areas, and its working groups. Note that other groups may also distribute working documents as Internet-Drafts. Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress." The list of current Internet-Drafts can be accessed at http://www.ietf.org/1id-abstracts.html The list of Internet-Draft Shadow Directories can be accessed at http://www.ietf.org/shadow.html Copyright and License Notice Copyright (c) 2013 IETF Trust and the persons identified as the document authors. All rights reserved. This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents Expires December 30, 2013 [Page 1] INTERNET DRAFT June 28, 2013 carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License. Table of Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 Use Case for Data Collection of ISP Network Management Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3 Security Considerations . . . . . . . . . . . . . . . . . . . . 6 4 IANA Considerations . . . . . . . . . . . . . . . . . . . . . . 6 5 References . . . . . . . . . . . . . . . . . . . . . . . . . . 6 5.1 Normative References . . . . . . . . . . . . . . . . . . . 6 5.2 Informative References . . . . . . . . . . . . . . . . . . . 6 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 6 Expires December 30, 2013 [Page 2] INTERNET DRAFT June 28, 2013 1 Introduction To support new services in network and provide better service quality, ISPs have to reconstruct or update their network constantly, which makes the network more and more complex and irregular. ISPs eager to have a comprehensive network management system to make the complex network in control to achieve real-time network performance acquisition, rapidly and accurately diagnose network fault with low cost, and SLAs of users satisfaction. Traditional network management systems consist of many different measurement panels, some of which are measurements in isolated network probes to report serious faults and some of which are measurements initiated by end users or dedicated network devices to locate the problem happening in their service paths or to calculate their performance of subscribed services. However, running network is always invisible to either operators or users. Through these measurements, traditional network management system could only interpret some scattered and fragmentary periods and paths while can't draw the whole running network picture for operators. It is also hard for them to deal with some difficult and complicated faults, such as sudden transient performance decrease which won't trigger any alarms. This document introduces a new use case supplementing the use cases described in [LMAP-USECASE]. It discusses measurements for a common platform which works for different usages including troubleshooting, performance understanding, quality evaluation and network adjusting. 2 Terminology The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119 [RFC2119]. 3 Use Case for Data Collection of ISP Network Management Systems Network data collection is essential for effective network management. Precise data collection and reasonable data analysis and processing could provide necessary information for network performance management, fault management and quality of service monitoring. Lmap could be used in network data collection of network management system as a universal method to sufficiently manage data collecting activities. Rather, lmap could be implemented in an intelligent network management platform to obtain accurate sample data for creating visualized simulation network which simulates the real one. By doing this, the intelligent visualized platform could draw a whole running network picture for ISPs and provide the better capability for troubleshooting, monitoring performance, and even Expires December 30, 2013 [Page 3] INTERNET DRAFT June 28, 2013 network planning instead of doing all kinds of measurements in the real networks. This intelligent network management platform using lmap constantly collects traffic information and device states, e.g. queue information, from the network, analyses them, and creates some sampled snapshots of the real network. By some simulation algorithms, these sampled snapshots could form a simulation network which simulates the real network accurately if sufficient information is retrieved. Usages, such as troubleshooting and quality monitoring, could run on the simulation network instead of using extra probes and measurement in the real network. For example, to learn what happened at a certain past time, ISPs could simply use the sampled snapshots of that time to infer the whole network curve of that past specific period. By further investigating, this intelligent platform could easily provide the ability to find the failure reason or discover the pattern of some complicated faults, e.g., one specific router has overflowed queues, which causes network congests. In this case, MAs are network devices constituting the whole network. Only passive measurements are needed since MAs just monitor the device states, network status and traffic information. No extra payload will be added to the existing network. Network-specific parameters are enough for this usage. Service-specific parameters will only be required in the simulation network, which is not in the scope of LMAP. This use case must consider and alleviate the performance issues caused by sample frequency and heavy measurement results reporting. As we know that the higher the sample frequency is, the closer the sampled network curve is to real network curve. But too frequent sample times will increase the burden of MAs and exhaust the resources of network devices. To solve this problem, some mechanisms, for example, using lower sample frequency when data vary gently while increasing sample frequency when network data change dramatically, should be considered to adjust data collection frequency. Another concern is the MA implementation in the network. The "ideal" situation is MA in each device (e.g. routers, switches) of the network ISP wishes to manage. However it is not feasible because we could envision heavy measurement report traffic disrupting the normal network traffic in large scale case. So only those network nodes arranged in a crisscross pattern and those important network devices to ISPs should be considered. Due to the high requirements of precise data collection and large scale environment, traditional protocols like SNMP are insufficient to do this kind of work in such huge and continuously expanding networks because of their constraints, e.g., producing plenty of management data which may causes serious traffic congestions, and incoordination among different network devices from different Expires December 30, 2013 [Page 4] INTERNET DRAFT June 28, 2013 vendors. Normally, usages in other use cases described in [LMAP-USECASE] are designed with some corresponding specific measurements. For example, measurements for identifying network problems, or measurements for evaluation the quality experienced by end users, etc. While different from other use cases, lmap used in this case is not dedicated for certain services, usages or end users. Instead, it is used to create a common and universal network management platform for all kinds of usages required by ISPs, including troubleshooting, performance evaluation, and other functions. The characteristics of large scale measurements emerging from this use case: 1. Passive measurements are needed while active ones aren't. 2. Network device states are also required as well as specific network performance parameters. Metrics of upper layer 3 are not. 3. The data collection frequency of passive measurements could be adjusted adaptively. 4. Results from the tests should not be averaged. 5. Regular scheduled tests are necessary. Expires December 30, 2013 [Page 5] INTERNET DRAFT June 28, 2013 3 Security Considerations TBD 4 IANA Considerations TBD 5 References 5.1 Normative References [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate Requirement Levels", BCP 14, RFC 2119, March 1997. [LMAP-USECASE] Linsner, M., "Large-Scale Broadband Measurement Use Cases", draft-linsner-lmap-user-cases-02, February, 2013 5.2 Informative References [LMAP-REQ] Schulzrinne, H., "Large-Scale Broadband Performance: Use Cases, Architecture and Protocol Requirements", draft- schulzrinne-lmap-requirements, September, 2012 Authors' Addresses Rachel Huang Huawei Technologies Co., Ltd. 101 Software, Yuhua District Nanjing, Jiangsu, 210012 P.R.China EMail: rachel.huang@huawei.com Expires December 30, 2013 [Page 6]