The National Laboratory for Applied Network Research (NLANR). http: //moat.nlanr.net/IPaddrocc/.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....38 40 Fraction of Collided Packets Prefix Length (in bytes) WAN LAN Figure 3: The fraction of packets that collide as a function of prefix length. The WAN trace represents 985,150 packets (with 5,801 duplicates removed) collected on July 20, 2000 at the University of Florida OC 3 gateway [14]. The LAN trace consists of one million packets (317 duplicates removed) observed on an Ethernet segment at the MIT Lab for Computer Science. wide area with identical prefixes indicates that packets with matching prefix lengths of 22 and 23 bytes are ICMP Time Exceeded error packets with the IP ....

NATIONAL LABORATORY FOR APPLIED NETWORK RESEARCH (NLANR). Network traffic packet header traces. http:// moat.nlanr.net/Traces/Traces/20000720/FLA964095596. crl.enc.

....simulate a model of a router with multiple NPRs and line cards. For router input, we have used pre generated traffic traces. Each trace corresponds to network traffic received at one line card. The parameters for generating the traces were approximated from OC 3 traces statistics gathered in [1] [15] and [22] and approximated to OC 192 speed by shortening the time intervals proportionally, i.e. 1 second of the monitored OC 3 traffic corresponds to 15 ms in our OC 192 like traces. The following parameters characterize the traces: Number of flows existent in a time interval a discrete ....

....[1] we have used exponential distribution with mean 4 to generate the individual flow lengths. Identifier vector values for the distribution of identifier vector values, we have approximated a typical distribution of IP source and destination addresses in networking traffic, as described in [15]. The prevalence of class C addresses, which occupy a relatively small portion of the address space (12.5 ) and yet account for approximately 65 of the packets in network traffic, led us to consider a normal distribution of identifier vectors within a 32 bit integer space, with parameters fitted ....

[Article contains additional citation context not shown here]

National Laboratory for Applied Network Research (NLANR). WAN traffic distribution by address size, Fix-West trace, http://www.nlanr.net/NA/Learn/Class, May 1997.

....to simulate a model of a router with multiple NPRs and line cards. For router input, we have used pre generated traffic traces. Each trace corresponds to network traffic received at one line card. The parameters for generating the traces were approximated from OC 3 traces statistics gathered in [1], 15] and [22] and approximated to OC 192 speed by shortening the time intervals proportionally, i.e. 1 second of the monitored OC 3 traffic corresponds to 15 ms in our OC 192 like traces. The following parameters characterize the traces: Number of flows existent in a time interval a ....

....(increase or decrease) in the number of packets is correlated with the direction of change in the number of flows (number of packets grows when number of packets grows and vice versa) as shown in [22] to hold. Flow length the amount of packets in a flow. Based on [22] and on the analysis of [1], we have used exponential distribution with mean 4 to generate the individual flow lengths. Identifier vector values for the distribution of identifier vector values, we have approximated a typical distribution of IP source and destination addresses in networking traffic, as described in ....

National Laboratory for Applied Network Research (NLANR) AIX - MAE West interconnection at NASA Ames OC-3 trace. http://moat.nlanr.net/PMA/, March 19, 2000.

....CHAPTER III SIMULATION RESULTS In our research, we use Internet packet traces obtained from http: pma.nlanr.net website to test the effectiveness of our proposed scheme. These measurements were carried out using a PMA monitor which is connected between NYSERnet s router and Buffalo s router [24] (OC3 Packet over Sonet (PoS) via a fiber tap. The traces provide information regarding the source, destination IP addresses, the packet size and the time stamps with a resolution of 1 s. The OC3 link speed is 155 Mbps. First, we provide information about the traffic on which we simulated the ....

"National Laboratory for Applied Network Research (NLANR)- BUF site infor- mation," http://pma.nlanr.net/PMA/Sites/BUF.html, accessed on 01/05/2002.

....SACRIO likely to be in real world when a router may observe thousands of flows How big a cache is likely to be required for SACRIO to be effective How scalable is SACRIO To answer these questions, we have looked at traces of real world networks. We studied traces available publicly from NLANR [14] to gain an understanding of the scalability of SACRIO. We collected six traces representing traffic at different times on different days. We reran our simulation with the traces to see the effectiveness of a small amount of state. The traces are all collected over a period of about 95 seconds of ....

National Laboratory for Applied Network Research (NLANR), Wide area Network Traces, available from http://moat.nlanr.net/Traces, June 2000.

....simulate a model of a router with multiple NPUs and line cards. For router input, we have used pre generated traffic traces. Each trace corresponds to network traffic received at one line card. The parameters for generating the traces were approximated from OC 3 traces statistics gathered in [1] [15] and [22] and approximated to OC 192 speed by shortening the time intervals proportionally, i.e. 1 second of the monitored OC 3 traffic corresponds to 15 ms in our OC 192 like traces. Note that this transformation is a simplification from reality, since the scaled traces would differ not only ....

....[1] we have used exponential distribution with mean 4 to generate the individual flow lengths. Identifier vector values for the distribution of identifier vector values, we have approximated a typical distribution of IP source and destination addresses in networking traffic, as described in [15]. The prevalence of class C addresses, which occupy a relatively small portion of the address space (12.5 ) and yet account for approximately 65 of the packets in network traffic, led us to consider a normal distribution of identifier vectors within a 32 bit integer space, with parameters fitted ....

[Article contains additional citation context not shown here]

National Laboratory for Applied Network Research (NLANR). WAN traffic distribution by address size, Fix-West trace, http://www.nlanr.net/NA/Learn/Class, May 1997.

....to simulate a model of a router with multiple NPUs and line cards. For router input, we have used pre generated traffic traces. Each trace corresponds to network traffic received at one line card. The parameters for generating the traces were approximated from OC 3 traces statistics gathered in [1], 15] and [22] and approximated to OC 192 speed by shortening the time intervals proportionally, i.e. 1 second of the monitored OC 3 traffic corresponds to 15 ms in our OC 192 like traces. Note that this transformation is a simplification from reality, since the scaled traces would differ not ....

....(increase or decrease) in the number of packets is correlated with the direction of change in the number of flows (number of packets grows when number of packets grows and vice versa) as shown in [22] to hold. Flow length the amount of packets in a flow. Based on [22] and on the analysis of [1], we have used exponential distribution with mean 4 to generate the individual flow lengths. Identifier vector values for the distribution of identifier vector values, we have approximated a typical distribution of IP source and destination addresses in networking traffic, as described in ....

[Article contains additional citation context not shown here]

National Laboratory for Applied Network Research (NLANR) AIX - MAE West interconnection at NASA Ames OC-3 trace. http://moat.nlanr.net/PMA/, March 19, 2000.

....information that is not completely accurate. This lack of accuracy is mainly due to the fact that mapping the Internet topology is a very challenging task [8, 17] At the Autonomous System (AS) level, available information is richer because it can be obtained or inferred from BGP tables [12, 7]. In contrast, accurate router level topological information is hard to obtain and until now inferring router level connectivity has been done by using traceroute or traceroute like probing mechanisms [8, 4] Identifying the actual fundamental properties of topologies at the router level is still ....

....hand written by the user or automatically generated by BRITE s GUI (see Section 3.8) BRITE provides the capability of importing topologies (2) generated by other topology generators (GT ITM [3] Inet [9] Tiers [5] BRITE 1. 0 [11] or topological data gathered directly from the Internet (NLANR [12], Skitter [4] Note that we include BRITE BRITE GT ITM NLANR Skitter Path Length Distribution Etc Outdegree Rank Outdegree Frequency Analysis BRITE Engine (4) MinHop Paths Distribution (2) File (Parameters) Configuration BRITE Engine Generation (1) BRITE NS SSF (3) ....

National Laboratory for Applied Network Research (NLANR). URL = http://moat.nlanr.net/rawdata/.

....We found several reasons to extend that study. Since large sets were not a consideration of Myers study, our measurement study extends Myers work by increasing the number of servers from 47 to 193. Second, considerable changes in the Internet warrant a new study. By examining logs kept by NLANR [4] of BGP tables (which stopped in January 2000) we can extrapolate that there has been a 40 increase in the number of Internet domains and a 49 increase in the number of domain interconnections since Myers study was conducted in April 1998 (see Figure 1) Third, Myers study involved ....

.... policies based on either the random selection of a server or minimizing the mean of ve roundtrip measurements consisDate Domains Interconnections Nov 8, 1997 3015 5539 Apr 15, 1998 3579 7011 Mar 1, 1999 4738 9812 Jan 2, 2000 6474 13895 Figure 1: Changes in the Internet Topology since 1997 [4]. tently outperform policies minimizing network hops between client and server. Similar work has been completed by others, including: comparing ping, hop count, and probabilistic policies [13] comparing AS hops, router hops and roundtrip time [10] comparing client probing and serving push ....

The National Laboratory for Applied Network Research (NLANR). http://www.moat.nlanr.net/Routing/rawdata.

....We found several reasons to extend that study. Since large sets were not a consideration of Myers study, our measurement study extends Myers work by increasing the number of servers from 20 to 193. Second, considerable changes in the Internet warrant a new study. By examining logs kept by NLANR [4] of BGP tables (which stopped in January 2000) we can extrapolate that there has been a 40 increase in the number of Internet domains and Date Domains Interconnections Nov 8, 1997 3015 5539 Apr 15, 1998 3579 7011 Mar 1, 1999 4738 9812 Jan 2, 2000 6474 13895 Figure 1: Changes in the Internet ....

.... stopped in January 2000) we can extrapolate that there has been a 40 increase in the number of Internet domains and Date Domains Interconnections Nov 8, 1997 3015 5539 Apr 15, 1998 3579 7011 Mar 1, 1999 4738 9812 Jan 2, 2000 6474 13895 Figure 1: Changes in the Internet Topology since 1997 [4]. a 49 increase in the number of domain interconnections since Myers study was conducted in April 1998 (see Figure 1) Third, Myers study involved characterizing the performance of servers based on transfer times alone, and did not measure the relative importance of client metrics in ....

The National Laboratory for Applied Network Research (NLANR). http://www.moat.nlanr.net/Routing/rawdata.

....w(t, #) # f(t, #) # # 1) 26) Thus, these figures imply the time translation invariance of ISGF, at least in the asymptotic region, # # 1. A. 2 Access Logs from Cache Server of NLANR As an example of backbone networks of the Internet, we investigate access logs of a cache server at NLANR [17]. Logs are for July 30, 1999. Figure 5 (c) shows the relationship between working set size, w(t, #) and the number of accesses, # , for the destination IP addresses in log log scale. Each line indicates w(t, #) for t =0, 2 10 5 , and 4 10 5 , respectively. Similarly, Fig. 5 ....

National laboratory for applied network research (NLANR)/ Measurement and operations analysis team (MOAT), http://moat.nlanr.net .

....information that is not completely accurate. Such lack of accuracy is mainly due to the fact that mapping the Internet topology is a very challenging task [11, 23] At the Autonomous System (AS) level, available information is richer because it can be obtained or inferred from BGP tables [17, 10]. In contrast, accurate router level topological information is hard to obtain and until now inferring router level connectivity has been done by using traceroute or traceroute like probing mechanisms [11, 6] Identifying the actual fundamental properties of topologies at the router level is still ....

....written by the user or automatically generated by BRITE s GUI (described in Section 4) BRITE provides the capability of importing topologies (2) generated by other topology generators (GT ITM [5] Inet [13] Tiers [8] BRITE 1. 0 [16] or topological data gathered directly from the Internet (NLANR [17], Skitter [6] Note that we include BRITE in the imported file formats, because it is possible to generate topologies using BRITE and then reusing them to generate other topologies by combining them with BRITE models or other imported formats. In the current distribution BRITE produces a ....

National Laboratory for Applied Network Research (NLANR). http://moat.nlanr.net/rawdata/.

....perceived at the proxy by up to a factor of 12. Related work is described in Section 6, and we conclude in Section 7. 2 Workload Characterization We examined the traces from two proxies (see Table 2) a proxy from a major national ISP, and the sv.cache.nlanr.net proxy managed by NLANR [19]. ISP NLANR Proxy Type Lowest level Upper Level Clients United States Asia and Pacific Dates 1 16 99 2 10 99 11 11 99 12 26 99 # requests 1.88 million day 1.28 million day Figure 2: Details about the proxy traces we studied. 2.1 Load prediction We first measured the load of traffic targeted to ....

National Laboratory for Applied Network Research (NLANR). A distributed testbed for national information provisioning. http://ircache.nlanr.net/Cache/.

....the packet at the destination machine. The accuracy of the delay measurements is within 50 microseconds. The test packets are 40 byte UDP packets, sent as a Poisson [40] stream, usually at an average rate of two packets per second. AMP The National Laboratory for Applied Network Research (NLANR; [11]) has a project called Active Measurement Program (AMP; 12] This project undertakes site tosite measurement across the vBNS [10] infrastructure. Around 30 AMP partici 1 A union of some U.S. universities. 12 pating sites are making connectivity, loss and round trip time measurements and ....

National Laboratory for Applied Network Research (NLANR) web site. http: //www.nlanr.net/, June 1999.

....di erent formats, but what we used is the tsh format ( tsh) with 44 byte xed records per packet, which gives us the transport and network layer protocol headers. The format is given in Appendix B. In addition to the raw traces, NLANR also o ers summary and analysis information for each trace. [16] Results available from NLANR include analysis of an aggregation point data, analysis of packets, bit volume, and host activity on a link, TCP ags, trac by protocol, and the behavior of sequences of packets or packet run lengths. 7 C. Trac Modeling The Notion of Flows For the purpose of ....

National Laboratory for Applied Network Research (NLANR), measurement and analysis team (MOAT), \Datacube" [Online]. Available: http://moat.nlanr.net/Datacube, accessed in September 2000.

....in collecting and analyzing the global Internet topology and traffic information, as well as many informative WWW sites. We describe the information sources we have used in our work, as well as several other related sources. 4. 1 NLANR The National Laboratory for Applied Network Research [6] has as its primary goal to provide technical, engineering, and traffic analysis support of high performance research networks. NLANR works towards a better understanding of the global Internet services and metrics. It collects and analyses data obtained through a range of passive and active ....

National Laboratory for Applied Network Research (NLANR). Online. .TTP://MO^T.NL^NR.NE/.

....are SNMP based network management tools, tcpdump [15] NetFlow [4] or DAG card based systems [21] Active measurement methods inject measurement traffic into the network and therefore interfere with operational traffic. Active measurement systems include e.g. NIMI [18] Surveyor [13] and AMP [17]. The AQUILA QoS measurement approach integrates both active and passive measurement methods. A third kind of classification of measurement methods is the distinction between aggregation based measurement and sampling based measurement [8] Aggregation based measurement methods collect and ....

National Laboratory for Applied Network Research (NLANR). Active Measurement Project (AMP), August 2001. Information available at: http://watt.nlanr.net.

....mechanism used by NLANR is called OC3MON, and is described in [14] NLANR uses, in parallel, other mechanisms, like active measurements (probing) and SNMP information to verify the accuracy of the collected information. The traces used in this thesis can be retrieved from NLANR s website [15]. Perl scripts are also provided on the website which can be used to decode the packet header into human readable form and or decrypt the encrypted headers. The traces provided are from various organizations participating in the program. The traces are keyed depending on the sources. Appendix A ....

National Laboratory for Applied Network Research (NLANR), measurement and analysis team (MOAT), \Network trac packet header traces" [Online]. Available: http://moat.nlanr.net/Traces, accessed in September 2000.

....degrade the service nor will it hinder its scalability. 6.1 Simulation Environment We implemented a custom simulator to study the performance of GIA. For our simulation topology, we use a set of snapshots of the Internet inter domain topology generated by NLANR based on the BGP routing tables [24]. Complete information about the simulation topology is provided in Appendix A1. In the absence of an anycast service from the current Internet, there is no data about the usage or characteristics of anycast groups. Therefore, we had to use some assumptions to carry out our simulations. We ....

....provider propagates a search received from another top level provider, it generates many redundant messages, which the above rules prune. We believe these rules are fairly simple and can be exercised easily by any ISP. The information necessary to support these rules is publicly available at [24]. Also, it can be gleaned locally at any ISP from the BGP routing tables. Note that this model tends to overestimate the number of search messages because transit domains do not exercise full control over the scope of a search. Finally, our simulator does not have the ability to simulate learning ....

[Article contains additional citation context not shown here]

The National Laboratory for Applied Network Research (NLANR), http://www.moat.nlanr.net/AS/.

No context found.

The National Laboratory for Applied Network Research (NLANR). http: //moat.nlanr.net/IPaddrocc/.

No context found.

National Laboratory for Applied Network Research (NLANR). Raw routing table information, 2001. http://moat.nlanr.net/Routing/ rawdata/.

No context found.

National Laboratory for Applied Network Research (NLANR), Traces collected in June 2000, http://pma.nlanr.net/pma/.

No context found.

National Laboratory for Applied Network Research (1997). [Online]. Available: http://moat.nlanr.net/Routing/rawdata

No context found.

National Laboratory for Applied Network Research (NLANR), races collected in June 2000.Av0.T[N; from <http://pma.nlanr.net/pma/>.

No context found.

National Laboratory for Applied Network Research (NLANR), Traces collected in June 2000, http://pma.nlanr.net/pma/.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC