Milestone MS3.7.5: Report on Passive Monitoring Pilot

Size: px
Start display at page:

Download "Milestone MS3.7.5: Report on Passive Monitoring Pilot"

Transcription

1 Milestone MS3.7.5: Report on Passive Monitoring Pilot Milestone MS Contractual Date: 31/05/08 Actual Date: 21/08/08 Contract Number: Instrument type: Integrated Infrastructure Initiative (I3) Activity: SA3 Work Item: 7 Nature of Deliverable: R (Report) Dissemination Level PU (Public) Lead Partner CESNET Document Code Authors: Sven Ubik (CESNET), Vladimir Smotlacha (CESNET), Szymon Trocha (PSNC), Simon Leinen (SWITCH), Vedrin Jeliazkov (BREN), Ales Friedl (CESNET), Gina Kramer (DANTE) Abstract This document describes the work carried out during the deployment of the Service Activity (SA) 3 passive monitoring pilot in the GÉANT2 () network. This pilot was conducted to follow up the development of passive network monitoring tools in the Joint Research Activity (JRA) 1 activity, in order to test these tools in a large-scale real network environment, gain experience in using these tools, gather user feedback and draw conclusions about the tool s usability and usefulness.

2 Table of Contents 0 Executive Summary vi 1 Introduction 1 2 Deployed Hardware 3 3 Developed and Deployed Software Architecture 8 4 Passive Monitoring Applications ABW Operation Usage Examples Difference from Related Tools Summary Tbwtools Operation Usage Examples Difference from Related Tools Summary Packetloss Operation Usage Examples Difference from Related Tools Summary Burst Operation Usage Examples Difference from Related Tools Summary 37 Document Code:

3 Contents 5 Service Applications Perfmon Operation Usage Examples Difference from Related Tools Summary Servmon Operation Usage Examples Difference from Related Tools Summary 45 6 The Relation of Applications in the SA3 Passive Monitoring Pilot to perfsonar 46 7 Challenges 48 8 Conclusions and Future Work 50 9 References Acronyms 53 Table of Figures Figure 3.1: DiMAPI architecture 8 Figure 4.1: Integrated access to all monitoring applications. 11 Figure 4.2: Map-based user interface of the ABW application. 13 Figure 4.3: ABW user interface. 14 Figure 4.4: Link load plotted with maximum values (left) and without maximum values (right). 15 Figure 4.5: Examples of link load in L3/L4 protocols (left) and application protocols (right) monitored by ABW. 16 Figure 4.6: SNMP link load monitoring. 17 Figure 4.7: Most detailed link load in perfsonar UI (left) and the ABW application (right). 18 Figure 4.8: Tbwtools architecture. 20 Figure 4.9: Tbwtools user interface. 22 Document Code:

4 Contents Figure 4.10: A test connection summary and performance characteristics. 23 Figure 4.11: A test connection summary and performance characteristics. 24 Figure 4.12: Packetloss application principle. 27 Figure 4.13: The main page of the Packetloss user interface. 29 Figure 4.14: Packetloss status and peer statistics. 30 Figure 4.15: Expired and matched flows from SWITCH to PIONIER. 31 Figure 4.16: Packet loss count and rate or real traffic. 31 Figure 4.17: Hades active loss monitoring. 32 Figure 4.18: User interface of the Burst application. 35 Figure 4.19: Distribution and cumulative distribution functions of frame burst sizes in bytes. 36 Figure 4.20: Time evolution of distribution and cumulative distribution of frame burst sizes in bytes as 3d graphs. 36 Figure 5.1: User interface of Perfmon application. 40 Figure 5.2: History of test events found by Perfmon. 40 Figure 5.3: Table of component states provided by Perfmon application. 41 Figure 5.4: Map-based user interface of the Servmon application. 43 Figure 5.5: Example of CPU usage and load monitoring. 44 Figure 5.6: Example of shared memory usage and DAG card drop count monitoring. 44 Document Code:

5 0 Executive Summary Network monitoring plays an important role in network planning, analysing network usage trends and resolving network usage issues. It also provides an information base for researchers of networking technologies. The JRA1 and SA3 activities have respectively developed and deployed the perfsonar framework in the network. perfsonar primarily uses active monitoring tools to measure network characteristics, such as oneway delay or TCP/UDP throughput. However, many interesting characteristics of real traffic in the network are inherent properties of that traffic and cannot be measured by test packets (for example, link load dynamics, real packet loss, composition of protocol traffic and applications, and trends in their usage). Passive monitoring is in principle much more powerful than active monitoring, because it can calculate any possible characteristics of real traffic. SA3 have developed several passive monitoring applications to address this, and deployed them for evaluation on the -NREN (National Research and Education Networks) links of four partners (ACAD [ACAD], CESNET [CESNET], PSNC [PSNC] and SWITCH [SWITCH]). This is the first phase of passive monitoring deployment in the network 1 : ABW (Available Bandwidth) - Monitors link load in short intervals to detect peak load. It also provides information about protocols in different layers that comprise link load, including multicast, IPv6 (Internet Protocol version 6) and some application protocols that use dynamic ports. Tbwtools - Tests TCP throughput between monitoring stations in the network and the user's PC, monitors this test connection by various techniques and provides feedback about connection health to assist in debugging low throughput. Packetloss - Monitors real packet loss of real network traffic, which is generally unrelated to packet loss that can be measured by test packets. Burst - Quantifies the burstiness of traffic independently from time intervals normally used to calculate average load. Perfmon and Servmon - Service applications that support the reliable operation of monitoring applications. 1 The fifth partner LITNET originally included in the pilot withdrew because their planned upgrade of link from 1 Gb/s to 10 Gb/s was delayed and it was not considered economical to buy a 1 Gb/s monitoring card that would need to be replaced soon. Document Code: v

6 Executive Summary The deployment of the pilot proved to be a valuable experience that made it possible to enhance applications to cope with large volumes of real traffic, to test the behaviour of multiple applications simultaneously sharing monitoring stations and to gather feedback from trial users. Multiple technical challenges were successfully tackled (see Challenges on page 48). The passive monitoring pilot deployment significantly extended the portfolio of useful monitoring applications in the network. It pioneered concepts that were not previously available from perfsonar, such as integrated access to all monitoring applications and map-based user interfaces. However, further steps should be taken in the development and deployment of passive monitoring in the and GN3 networks. Passive monitoring services will become increasingly useful as they are deployed across and beyond most of /GN3-NREN links in participating NRENs. New applications will also provide additional benefits, such as Service Level Agreement (SLA) verification for specified user traffic in projects that use the network for more detailed traffic classification in order to understand network usage trends. Document Code: vi

7 1 Introduction Network monitoring plays an important role in network planning, analysing network usage trends and resolving network usage issues. It also provides an information base for researchers of networking technologies. Most network monitoring methods can be grouped into: Active monitoring. Passive monitoring. Network infrastructure monitoring. In active monitoring, test packets are sent across the network to measure various performance characteristics. Active monitoring is a convenient way to measure, for instance, one-way delay over a network. However, the results of active monitoring are only truly valid for the test packets, not for real user traffic. In passive monitoring, test packets are not sent across the network. Instead the characteristics of real network traffic are observed and analysed. Many important network performance characteristics are inherent properties of real traffic, and can therefore only be obtained from passive monitoring. In principle, passive monitoring is a more powerful approach than using test packets, but it is also more computing-intensive, as large volumes of data need to be analysed in real time. It therefore requires relatively expensive packet capture hardware (monitoring cards). The following are examples of performance characteristics that can only be obtained using passive monitoring or a combination of passive and active monitoring: Link load in a wide range of time scales. Protocols and applications used in the network, and trends in their usage. Real packet loss experienced by users. Detailed feedback about connection health for performance debugging. Various kinds of anomalies. Various kinds of security problems. Passive monitoring also has important properties for the quality of monitoring and network operation: It does not affect user traffic. It can run continuously. It identifies characteristics of real user traffic. Document Code:

8 Introduction Network infrastructure monitoring uses information obtained from network components, such as routers and switches. The two most common protocols of network infrastructure monitoring are SNMP and Netflow. Both protocols are commonly used in reliable monitoring which provides useful statistics about network traffic. However, for both protocols only information provided by specific router and switch makes can be obtained. Common limitations are that SNMP provides only high-level statistics about network traffic in coarse time granularity (such as total volume), and that Netflow records include a fixed set of information fields. Netflow records can also be obtained from passive monitoring, in which case they can be extended with additional information fields. The perfsonar system developed in the JRA1 activity is almost exclusively based on active monitoring. Therefore, it was necessary to extend the range of monitoring methods in the network with passive monitoring to provide key performance characteristics inherent to real user traffic, which could not be obtained from active monitoring. The development of passive monitoring tools and infrastructure in the JRA1 activity was the responsibility of CESNET. As the tools matured enough to be deployed in the network, CESNET passed their experience to the SA3 activity and realised a pilot deployment of passive monitoring in four project partners as described in this deliverable. The pilot aimed to extend the portfolio of monitoring services in the network with additional useful characteristics of real network traffic. Document Code:

9 2 Deployed Hardware To capture and process traffic at multi-gigabit rates, it is necessary to use monitoring cards rather than regular Ethernet adapters. Monitoring cards provide two primary advantages: They transfer packets from the network to the memory of a host PC much more effectively, with almost zero CPU load, leaving CPU for processing of packets. Extensive comparative performance measurements of several monitoring cards and Ethernet adapters [TNC08] confirmed that a modern PC can be used for complex monitoring at multi-gigabit rates, provided that a monitoring card is used to capture packets. They usually provide some hardware acceleration, such as filtering, simple classification and packet statistics. For the pilot deployment, several network monitoring options were explored. The Force10 P-series box, which is primarily designed for 10 Gb/s firewalls, was found not to be suitable for the required monitoring purposes. Napatech s 10 Gb/s monitoring cards were not commercially available. The only commercially available monitoring cards for 10 Gb/s links at the time of procuring hardware for the pilot were Endace DAG cards. Therefore, suitable types of these were used for the -NREN links of pilot partners. The deployed infrastructure consists of 7 monitoring stations and one central station for application support and user interfaces 2. Based on previous experiences and performance tests, Supermicro servers were used in the pilot. The hardware of 5 monitoring stations deployed in the pilot is configured as follows: Supermicro X7DB8+ mainboard 2x Xeon DualCore 3 GHz 2x 73 GB SCSI disk 2 GB FB-DDR2 667 MHz RAM 2 Although two monitoring PCs were installed in CESNET as part of the JRA1 activity before the pilot, the software on these two monitoring stations had to be reinstalled. The other 5 monitoring stations were deployed as part of this pilot. Document Code:

10 Deployed Hardware CESNET monitoring stations: Intel SE7520 mainboard 1x Xeon 3 GHz (single core) 73 GB SCSI disk 1 GB RAM Central station: Supermicro mainboard Xeon DualCore 2 GHz 150 GB SCSI disk 2 GB RAM The monitoring cards are connected to the -NREN links by optical splitters. An exception is the ACAD location, where a mirroring port on a router is used to supply traffic to the monitoring card. This allows the use of twisted-pair transceivers which for logistics reasons are more convenient at this location. The following table summarises all 7 monitoring stations and one central station in CESNET: Partner Locality Monitoring station Monitoring card Monitored link ACAD Sofia, Bulgaria sa3-pm5.acad.bg DAG4.5G4 (4x 1 Gb ethernet) ACAD to, to ACAD CESNET Prague, Czech Republic jra1-1.cesnet.cz DAG6.2S (OC-192/STM-64) CESNET2 to jra1-2.cesnet.cz DAG4.3 (2x 1 Gb ethernet) CESNET2 to to CESNET2 sa3-pm.cesnet.cz, alias perfmon.geant2.net (central station) -- PSNC Poznan, Poland pm1.passive.pionier.net.pl DAG6.2S (OC-192/STM-64) to PIONIER pm2.passive.pionier.net.pl DAG6.2S (OC-192/STM-64) PIONIER to SWITCH Geneve, Switzerland gn2-pm1.switch.ch DAG8.2X (10 Gb ethernet) SWITCH to gn2-pm2.switch.ch DAG8.2X (10 Gb ethernet) to SWITCH Table 2.1: Monitoring station summary. Document Code:

11 3 Developed and Deployed Software Although it was planned to use applications that had previously been developed by the JRA1 activity and the LOBSTER [LOBSTER] project, most applications needed significant improvements to provide more performance stability and better user interfaces. The applications need to run in a large real network environment with much higher traffic volumes than in the local networks where they were developed and initially tested. Two new service applications (Perfmon and Servmon) were developed to support the infrastructure. For details of necessary improvements to individual applications, see their respective sections in Passive Monitoring Applications on page 11 and Challenges on page 48. The following table summarises the passive monitoring applications that were deployed within the pilot: Application Provided network characteristics Added value Users ABW Provides the used capacity in a wide range of time resolutions including short load peaks. It also provides distribution of traffic into protocols and applications. Identifies short-term traffic dynamics that can affect added traffic and protocols and applications that use up capacity or behave differently (e.g., elastic vs. nonelastic traffic), including trends in their usage. These characteristics cannot be obtained from active monitoring or network infrastructure monitoring. PERT NOCs Research projects using the network. Advanced users. Tbwtools Provides TCP throughput test with detailed feedback about the test connection for performance debugging. Helps investigate why TCP connections have low throughput. Uses an integration of active, passive and end-host monitoring. PERT Research projects using the network. Packetloss Provides detailed statistics about packet loss of real user traffic. Provides information about packet loss of real user traffic, which is generally different from packet loss of test packets. Measurements confirmed that real user traffic experiences packet loss which is not detected by active monitoring. PERT NOCs Research projects using the network. Advanced users. Document Code:

12 Developed and Deployed Software Burst Provides the burstiness of real traffic. Helps understand traffic dynamics (useful for researchers) and to evaluate conditions expected by added user traffic. PERT NOCs Research projects using the network. Advanced users. Table 3.1: Passive monitoring applications deployed in the pilot. Notes: The originally planned TCMP (Tracefile Measurement Point) application could not be deployed, because it lacks authorisation necessary for this type of application and the original developer in ARNES left the project. CESNET are currently working on adding authorisation to TCMP and expect to complete this shortly after the pilot. A new Burst application was added to the pilot instead of TCMP. Several support applications and important software components were also developed and deployed during the pilot: Application Provided network characteristics Added value Users Perfmon Monitors the state of hardware and software components of monitoring stations, sends alerts and automatically repairs problems where possible. Checks the status of monitoring stations hardware and software components. Users of measurements results benefit from increased application reliability. Servmon Monitors hardware and software resources on monitoring stations, such as CPU usage, CPU load, allocated shared memory and packet drops reported by monitoring cards. Checks resources on monitoring stations. Users of measurements results benefit from increased application reliability. DiMAPI (middleware) Middleware to support the operation of all passive monitoring applications. Enables the development of portable monitoring applications and their concurrent operation in a heterogeneous environment with transparent hardware acceleration. Necessary infrastructure to support applications. Table 3.2: Support applications and software components deployed in the pilot. Document Code:

13 Developed and Deployed Software Notes: Perfmon and Servmon are service applications developed to support reliable operation of the pilot infrastructure. They can also be used outside of the pilot. DiMAPI (Distributed Monitoring Application Programmable Interface) is middleware, rather than an application, but it forms a part of indivisible application functionality. Document Code:

14 Developed and Deployed Software 3.1 Architecture The architecture of passive monitoring applications based on DiMAPI is shown in Figure 3.1. Packets are processed as follows (from the bottom of the picture upwards): Packets are copied from the network by an optical splitter or a mirroring port on a router or switch. Packets are captured by a network card, which can be (in principle) a regular Ethernet adapter or a monitoring card. Packets are then processed by DiMAPI middleware, which is divided into: The mapid and mapicommd daemons running on monitoring stations with monitoring cards distributed across the network. DiMAPI libraries, which are linked to applications that can run on the same or different PCs than the monitoring stations. Typically on application PCs in order to use their separate processing power. Figure 3.1: DiMAPI architecture Document Code:

15 Developed and Deployed Software The deployed infrastructure has the following particular properties: Monitoring cards (rather than regular Ethernet adapters) are used on all monitoring stations. The ABW and Burst applications run on monitoring stations. Their results are retrieved on demand and sent to the central station that provides the user interface. This ensures that results are not affected by potentially variable communication delay between DiMAPI daemons and libraries in a distributed configuration. The Packetloss application runs on a central station and uses remote communication between DiMAPI daemons and libraries. The Packetloss application requires this configuration to be able to correlate data obtained from multiple monitoring stations and calculate resulting packet loss characteristics. Variable delay is not a problem for this application and it also benefits from running in part on a separate PC from highly loaded monitoring stations. Tbwtools uses an integration of active, passive and end-host monitoring. It has a different architecture (see Document Code:

16 Developed and Deployed Software Tbwtools on page 19). Perfmon and Servmon service applications are also distributed between monitoring stations and the central station, but they do not process packets and do not use DiMAPI mechanism for communication. Document Code:

17 4 Passive Monitoring Applications This chapter describes the passive monitoring applications that were deployed as part of this pilot. An important feature of application deployments in the pilot is that all measurement results are conveniently accessible from a single, central web page [WEBPAGE] (see Figure 4.1). Figure 4.1: Integrated access to all monitoring applications. Document Code:

18 Passive Monitoring Applications The top of the page includes a clearly organised table that lists each application s name, the network characteristics provided for it and a link to its current measurements. Some applications have two user interfaces; a map-based interface for a quick overview of the current state of monitored components or characteristics, and a form to request detailed tables and graphs. The rest of the page includes further details for each application, including information about the type of traffic being monitored, whether monitoring is on demand or continuous, the monitoring time resolution and where documents describing the applications can be found. For convenience, information how to access perfsonar active monitoring applications from the same page is also included. 4.1 ABW Operation ABW is an application to monitor used capacity. It primarily provides the following benefits: It shows used capacity in a wide range of time resolutions (from one second to months). It shows the distribution of used capacity across protocols and applications including the ones using dynamic ports. The ABW application has been developed by the JRA1 activity. It can detect which protocols real traffic on a network link consists of at different OSI (Open Systems Interconnection) hierarchy layers, or monitor which user-specified subsets of traffic based on arbitrary L2 to L4 header fields are present on the link. It uses the trackflib library of DiMAPI to identify many commonly used protocols that use dynamically allocated ports, such as passive FTP (File Transfer Protocol), HTTP (Hypertext Transfer Protocol, which can use other ports than 80 and 443), Skype telephony or file-sharing protocols, such as BitTorrent, Gnutella or DC++. It also monitors the volume of multicast and IPv6 traffic as two interesting subsets of traffic. The volume of traffic (separately the number of packets and the number of bytes) for all configured protocols is measured every second and stored in an RRD (Round Robin Database) where the results are gradually collected. Upon user request, the application generates graphs for specified monitored links, protocols, time intervals and time resolutions (predefined or user-defined). The application is therefore split into two phases: continuous data acquisition and data presentation. The graphs also show the average and maximum link load for the same time period, but with a coarser time resolution. Only a standard web browser is needed, no Java or plugin support is required and, as the user interface is generated on the server, there is no need to download any executable to the client. Generated graphs have one of two possible layouts showing either unidirectional or bidirectional line load. For bidirectional line load, one direction is shown as positive values and the opposite direction as negative values. The selected graph layout depends on the number of ports on the monitoring card. When the card has only one port (such as 10 GE of OC-192 DAG cards), ABW uses unidirectional layout. If the card has two ports, each monitoring one direction, ABW uses bidirectional layout. Document Code:

19 Passive Monitoring Applications Usage Examples ABW provides a map-based and a from-based user interface. The map-based user interface gives a quick overview of the most recent load on all monitored links (see Figure 4.2). Each monitoring station is represented by a coloured circle divided into two halves. The left half shows input load and the right half shows output load, indicated by arrows next to the circle. The colour and the filledin area depend on the average load in the last 15 minutes. If the cursor is placed on a circle s input or output half, a popup box displays more detail about the latest link load. Figure 4.2: Map-based user interface of the ABW application. Document Code:

20 Passive Monitoring Applications If the box is clicked, the form-based user interface is displayed (see Figure 4.3). Users can also invoke this interface directly from the main page of passive monitoring applications. Figure 4.3: ABW user interface. The ABW application s user interface is very similar to that of other passive monitoring applications, making it easy for users to get oriented. Reasonable defaults are preset, so that the user can just click the "Generate graphs" button. Details of requested graphs can be specified using the following steps which are clearly indicated in the user interface. Document Code:

21 Passive Monitoring Applications Step 1: The user can select one or more types of graphs. Currently two types of graphs are available: distribution of L3/L4 protocols including multicast and IPv6 and distribution of application protocols. Step 2: The user can select one or more monitoring stations. Step 3: The user can select the time period and resolution for which link load and protocol distribution should be presented. Several predefined time periods and resolutions are available. Alternatively, the user can specify an arbitrary time period and resolution. Similarly to other applications, one resolution can be specified for the graphs main body and one for envelope lines showing average and maximum values. Step 4: The user can choose if the envelope line for maximum values should be included in the graph. Maximum values are calculated from the original data samples normally received in one second intervals. For graphs with long-term average and maximum aggregation lines, the maximum values can be much higher than values drawn in finer resolution, which then occupy only a small part of the graph and are less clear. The example graphs in Figure 4.4 illustrate this effect. The left graph shows link load with the maximum line, and the right graph shows the same measurement without the maximum line. Figure 4.4: Link load plotted with maximum values (left) and without maximum values (right). Document Code:

22 Passive Monitoring Applications The example graphs generated from the form-based user interface are shown in Figure 4.5. The upper two graphs are for the -CESNET link and the lower two graphs are for the -ACAD link. Figure 4.5: Examples of link load in L3/L4 protocols (left) and application protocols (right) monitored by ABW. In ACAD, traffic in both directions is merged in a mirroring port on a router, monitored together and shown in a unidirectional graph. The left graphs show distribution of L4 protocols and the volume of multicast and IPv6 traffic. The right graphs show distribution of application protocols. Different protocols are shown in different colours. Grey represents traffic whose application protocol could not be determined. All graphs show link load with 1-second time resolution. The light and dark blue lines show 30- second average and maximum values respectively. You can see that short-term load (1-second) is often significantly higher than long-term (30-second) averages. This phenomenon is even more pronounced in longer-term monitoring. Figure 4.4, for instance, shows the link load on the -CESNET link over a one-week period of with 20-minutes resolution and maximum values calculated from 1-second samples. Document Code:

23 Passive Monitoring Applications Difference from Related Tools The ABW application provides several advantages over other approaches to link load monitoring: When compared to SNMP monitoring, passive link load monitoring can provide results at any time scale. This enables the detection of short load peaks that are often much higher than long term averages. These peaks cannot be detected by SNMP monitoring, because routers normally update their MIB (Management Information Base) counters with varying delays of several seconds. For example, Figure 4.6 shows SNMP link load monitoring at 30-second intervals on the left and at 1-second intervals on the right. The 1-second monitoring is unusable because it is distorted by interference between counter update and reading rates. Figure 4.6: SNMP link load monitoring. Another advantage of passive link load monitoring is that it provides detailed information about protocols and applications that use network capacity including usage trends. This information can be used for network planning or traffic engineering. When compared to link load presentation in perfsonar UI (which uses SNMP monitoring), the ABW application can show link load for any time period with any time resolution, provided that requested data is still available and not aggregated. For instance, if a problem is discovered in the network, detailed load dynamics and protocol changes that happened several hours or days ago are available for analysis. ABW can also simultaneously show multiple graphs for different links on one web page for easy comparison, and it allows downloading produced graphs to files, which is not possible with perfsonar UI. Figure 4.7 shows the most detailed link load information available in perfsonar UI on the left (5-minute resolution, most recent), and in ABW on the right (1-second resolution, any selected time period). Document Code:

24 Passive Monitoring Applications Figure 4.7: Most detailed link load in perfsonar UI (left) and the ABW application (right) Summary The ABW application has been deployed on all -NREN links of partners participating in the SA3 pilot. This provides the following benefits: Continuous and 100% non-intrusive link load monitoring. Link load for any time period and in any time resolution, ranging from 1-second peaks to averages over months for evaluation of load dynamics in the latest or previous traffic. Distribution of load into L4 protocols and application protocols, including common protocols that use dynamic ports. Monitoring of multicast and IPv6 traffic. Simultaneous display of multiple links for easy comparison. Fast user interface for a standard web browser. Document Code:

25 Passive Monitoring Applications 4.2 Tbwtools Operation Tbwtools is an application that uses a combination of active monitoring, passive monitoring and end-host monitoring to test and analyse TCP throughput over an end-to-end path. The letters tbw stand for tcpdump, bulk and web100, because these are the three monitoring utilities combined in Tbwtools. The application is similar to iperf [IPERF] or BWCTL (Bandwidth Test Controller [BWCTL]) stress-type throughput tests, but in addition to these tools it also provides connection diagnostics. Tbwtools provide the following primary benefits: The Tbwtools infrastructure deployed in permits the testing of network paths all the way from / to the user's PC, which can run Windows or a Linux operating system. Tbwtools is automatically downloaded from the central web page to the user's PC and started. (A deployed BWCTL infrastructure only permits tests between fixed servers in the network, or from the user's PC with a Linux operating system, where BWCTL must be installed manually.) It provides detailed feedback about connection performance characteristics to help determine the reason why throughput was low. Tbwtools has been initially developed by the LOBSTER [LOBSTER] project. In 2006, it was presented to the JRA1 activity and it was agreed that the further development would be taken over by this activity. Since then the application has been significantly enhanced. The user interface has been improved, support for LS (Lookup Service) has been added, the installation has been simplified and NATs (Network Address Translations), which are often present in local networks, are now supported. Tbwtools consists of a web-based user interface (implemented in Java and started through Java Web Start) and a perfsonar-compatible Monitoring Point (MP). Using the web-based user interface is optional. It is also possible to communicate with the MP by exchanging XML messages using a perfsonar client. However, the user interface allows to make a test connection from a user's PC (if the interface is not used, this is only possible between two MPs), and it presents monitoring results in a graphical form. As Tbwtools also use active and end-host monitoring, it is based on a different architecture to other strictly passive tools (see Figure 4.8). Document Code:

26 Passive Monitoring Applications Figure 4.8: Tbwtools architecture. The Tbwtools operate as follows (see Figure 4.8): 1. Tbw MPs (Measurement Points) are installed on monitoring stations in the network. When each Tbw MP is started during monitoring station boot, it automatically registers its URL with the Lookup Service (LS). Each Tbw MP can belong to a named domain, such as "". 2. A user points their web browser to a web page from which a Java-based user interface is downloaded and started. This user interface sends a query to the LS to obtain URLs of available Tbw MPs. The query to the LS includes the name of the user's domain, such as "", which is obtained from the web page where the user interface was downloaded. This allows Tbw MPs from multiple domains to be registered on the same LS, and the correct set of Tbw MPs of the user's domain to be presented to the user. 3. A user makes a test connection between two MPs or between a selected MP and their own PC. As this test connection is in progress, it is also being monitored by several methods (passive packet capture, web100 and getsockopt() system call). 4. After the test connection finishes, various performance characteristics are calculated and presented to the user in a set of graphs. Document Code:

27 Passive Monitoring Applications Usage Examples The main Tbwtools user interface page is shown in Figure 4.9. The user actions are divided into a few simple steps clearly marked on the user interface. Step 1: Clicking the "Query LS" button queries the LS for a list of available TBw MPs. Step 2: Available Tbw MPs are automatically added to a drop-down list. The user can select one of them or manually enter the URL of another Tbw MP not registered with the LS. Step 3: The user can specify the hostname or IP address of the other end of a test connection. By default this is the user's PC IP address, which means that a test connection is made between a selected Tbw MP and the user's PC. By entering another hostname or IP address, a test connection between two Tbw MPs can be made. A test from / to the user's PC is possible because the Java-based user interface also includes a Tbw MP implementation. Step 4: The user can optionally choose: The direction in which test data is sent during the test connection (to test a network path in a selected direction). The direction in which the test connection is initiated. (If there is a firewall on one side, the test connection is more likely to path through when started from behind the firewall. Note that the TCP port can be specified in step 2.) Step 5: The user can optionally specify: The socket buffer size (when unspecified, operating system defaults are used). The message size sent in one send() call (default 64 kb), web100 sampling rate (default 100 µs). Note that the message size also affects sampling rate for socket options (particularly TCP_INFO) read by getsockopt(), because this call is made after each send() call. The test duration (default 5 s). Finally, the user can click the "Run test" button to make a monitored test connection. Document Code:

28 Passive Monitoring Applications Figure 4.9: Tbwtools user interface. After the test connection finishes, the user is presented with a summary of the connection and a choice of approximately 100 performance characteristics, which were monitored during the connection and can be plotted (see Figure 4.10). A summary includes information about the percentage of time that the connection throughput was limited by a sender, a receiver or a network. This is important initial information about the connection bottleneck s location. The performance characteristics are produced from three monitoring sources: tcpdump - Characteristics calculated from captured packet trace of a test connection, such as instantaneous throughput, flight size, sequence and acknowledgement numbers. bulk - Characteristics calculated from the TCP_INFO socket option, read by the getsockopt() system call after each write of test data by send() call. For example, sender and receiver ssthresh (slow start threshold) variables. Document Code:

29 Passive Monitoring Applications web100 - Characteristics obtained from extended Linux kernel monitoring of TCP connections supported by the web100 kernel patch. Figure 4.10: A test connection summary and performance characteristics. After the user selects characteristics and clicks the "Generate graph" button, the characteristics are calculated and plotted in graphs. A section of a web page with generated graphs is shown in Figure An important benefit provided by Tbwtools is that characteristics obtained from different types of monitoring are timecorrelated and presented together in a common user interface. Document Code:

30 Passive Monitoring Applications Figure 4.11: A test connection summary and performance characteristics. The nature of characteristics requires expert TCP knowledge to analyse results. If a user does now have this knowledge themselves, they can still make a test connection from / to their PC, obtain a data ID from the results and it to expert who can use Tbwtools to retrieve the results for analysis Difference from Related Tools The following tools are most closely related to Tbwtools: iperf - A widely-used stress-type TCP/UDP throughput test tool. A "bulk" component of Tbwtools, it provides a very similar stress-type test (for TCP only). However, Tbwtools also includes test connection monitoring, provides an initial bottleneck location summary and detailed performance characteristics to assist expert users in detailed connection analysis. BWCTL - A wrapper around iperf with added scheduling to enable multiple users to execute tests without interference, deployed in as part of JRA1 and SA3 activities. The same key difference as with iperf applies to BWCTL, which does not provide feedback why measured throughput was potentially low. Tbwtools adds monitoring and detailed feedback about the test connection. Moreover, the installed BWCTL infrastructure measures throughput only between fixed monitoring stations, while Tbwtools measure throughput all the way to / from the user's desktop (measurements between two monitoring stations are also possible). Document Code:

31 Passive Monitoring Applications NDT (Network Diagnostic Tool) - This tool is similar to Tbwtools in that it combines a stress-type TCP throughput test with web100 monitoring. Tbwtools also collects characteristics obtained from passive packet capture and from the TCP_INFO socket option, and presents them correlated in time in a central user interface Summary The Tbwtools application has been deployed on all -NREN links of partners participating in SA3 pilot. This provides the following benefits: On-demand TCP throughput tests between monitoring stations or between monitoring stations and user PCs. Built-in test connection monitoring based on passive packet capture (tcpdump) and end-host monitoring (web100 and getsockopt() ). Connection summary including the achieved throughput and the percentage of time that the connection throughput was limited by a sender, a receiver or a network. Computing and plotting of up to approximately 100 selected performance characteristics obtained from several monitoring sources correlated in time for easy comparison. User-accessible monitoring information enables both self-analysis and forwarding of data IDs for expert monitoring result retrieval and analysis. Able to pass through NATs and some firewalls. 4.3 Packetloss Operation Packetloss is a new application, which is being tested on a large scale in the network for the first time. Packet loss is an important characteristic that affects user experience when communication takes place over a network. As the following example illustrates, it is unfortunately not possible to measure realistic packet loss using active monitoring: If 10 test packets per second were sent between two end points (much more than in the active monitoring tool Hades deployed in ), it would take almost 3 hours to detect packet loss of 10-5 and more than a day to detect However, these are packet loss rates that still significantly affect TCP throughput over fast long-distance networks. Moreover, these calculations are valid for evenly distributed packet loss. When bursts of packet loss occur, which is a common case, it can take an even longer time to detect and realistically measure packet loss rate by test packets. Document Code:

32 Passive Monitoring Applications The problem can also be looked at from another perspective: If a burst of 10 test packets was sent and a loss period was caught, so that 5 out of 10 packets were lost, it would not be possible to determine the time period for which this 50% packet loss rate is valid. Packet loss is a property inherent to traffic in which it is experienced. Due to the volume and dynamics of real traffic, it is not possible to realistically capture its packet loss using test packets. No matter how many and in what patterns they are sent, the outcome is not related to the actual network behaviour experienced by user traffic. This is completely different from one-way delay, which can be realistically measured by test packets in active monitoring. Packetloss is a strictly passive monitoring application for measuring real packet loss experienced by real user traffic. It gathers records about completed flows from monitoring stations on edges of a network (such as NREN - interfaces) and calculates real packet loss of flows passing through the network. It does not require any configuration for the IP prefixes that pass through the edges of a network. The basic idea is to compare the number of packets reported in the same flow from multiple monitoring stations. If the number of packets is different, it is the real packet loss which happened inside the network. This idea is illustrated in Figure However, the implementation is rather complex for achieving required performance properties. The flow record is created on a monitoring station when a flow expires (or completes); that is when no packet of the flow arrives during a specified inactive period (the default is 30 seconds). Flow records are periodically retrieved by the Packetloss application which runs on the central station. A flow reported from multiple monitoring stations is considered the same if it has the same 5-tuple (source IP address, destination IP address, source port, destination port and layer 4 protocol) and if the timestamps of the first packet do not differ more than a specified time (the default is 500 milliseconds). The packets of one connection can go via different routes inside the network and will be counted correctly as long as they arrive at the same edge router. Packets can arrive at different edge routers in case of multicast or where multiple uplinks of an NREN to the network exist and load balancing allows a one single flow to be split into multiple uplinks (which is an unusual configuration). These cases are not supported by the Packetloss application. The Packetloss application was originally developed by the LOBSTER project. However, for the purposes of deployment in the network, its performance and user interface needed to be improved significantly, which resulted in an almost complete redesign of the application including related parts of DiMAPI. Document Code:

33 Passive Monitoring Applications Figure 4.12: Packetloss application principle Usage Examples The main Packetloss user interface page is shown in Figure Its use is similar to other applications. Some reasonable defaults are preset, so that the user can just click the "Generate graphs" button. However, the generated statistics and graphs can be customised using the following steps which are clearly marked in the user interface: Step 1: The user can select NREN pairs between which monitored packet loss is to be shown. On the left, an NREN with a monitoring station is selected. On the right, one or more other NRENs can be selected. Packet loss will be shown between the NREN on the left and all NRENs on the right. Document Code:

34 Passive Monitoring Applications Step 2: The user can select the time period and resolution for which packet loss is to be shown. Several predefined time periods and resolutions are available. Alternatively, a user can specify an arbitrary time period and resolution. Similarly to other applications, one resolution can be specified for the main body of the graphs and one for envelope lines showing average and maximum values. Step 3: The user can select one or more of four graph types: packet loss ratio, packet loss count, number of all expired (completed) flows and number of flows matched between NRENs. Step 4: The user can select one of four available graph sizes. Step 5: The user can generate graphs as draggable to easily access previous or following time periods. Step 6: The user can request a number of characteristics to be listed in tables. Finally, the user can click the "Generate graphs" button to generate tables with statistics in textual form. Although graphs are generated for the requested time period, the tables include statistics for the whole time that the application is running. Statistics can be cleared at any time by clicking the "Reset statistics" button. The user can also enter header filter specification in BPF (Berkeley Packet Filter) format in the field below the "Generate graphs" button. Clicking the "Restart packetloss" button, restarts the application with the specified header filter. Only packets that pass this filter are included in the following monitoring. This feature is useful for users who are interested in monitoring only a subset of traffic, or if the total traffic volume is higher than the amount that the Packetloss application can process. Document Code:

35 Passive Monitoring Applications Figure 4.13: The main page of the Packetloss user interface. Document Code:

36 Passive Monitoring Applications The main page of the Packetloss user interface also includes the status table and the peer statistics table at the bottom (see Figure 4.14). The status table shows valuable information including: The BPF string used for header filtering. By default all packets are monitored. The minimum number of packets in a flow to be processed. By default flows with at least two packets are monitored. A higher value means more restrictive filtering, which reduces the precision of measured packet loss. Ignoring flows of 1-packet size is useful for performance reasons and for better precision. If these flows were not ignored, 1-packet flows reported from one monitoring station would be counted to the total number of flows, but lost 1-packet flows would not be detected and thus not counted to packet loss statistics. The status of all monitoring stations and their monitoring cards, which can be connected, disconnected due to problem on the monitoring station or disabled manually. The peer statistics table shows the total number of matched and lost packets for all selected NRENs pairs. Matched packets exist where flows were detected as passing from one NREN to another NREN. Figure 4.14: Packetloss status and peer statistics. Document Code:

37 Passive Monitoring Applications The left part of Figure 4.15 provides an example of a graph that shows expired (completed) flows reported from two monitoring stations (SWITCH and PIONIER in this case), specifically from the monitoring cards that monitor traffic from one NREN to and from to the other NREN during a period of one week. Matched flows (flows that were flowing from the first NREN to the second NREN) are shown in the right part of Figure Figure 4.15: Expired and matched flows from SWITCH to PIONIER. Figure 4.16 shows the measured packet loss count and packet loss rate. Figure 4.16: Packet loss count and rate or real traffic. Document Code:

38 Passive Monitoring Applications For comparison, Figure 4.17 shows graphs from active packet loss monitoring by the Hades application between the same NRENs during the same time period. There are two Hades monitoring stations in PSNC, shown on the left and right graph respectively. Active monitoring did not detect any packet loss, which could be expected, due to completely different volume and dynamics of a few test packets compared to real traffic. Figure 4.17: Hades active loss monitoring Difference from Related Tools As outlined in Operation on page 25, packet loss is strictly bound to particular traffic and it is not possible to obtain it using other traffic. While test packets in active monitoring can be used to measure delay, which corresponds to delay experienced by real traffic, it is a different case with packet loss. Packet loss provided by active monitoring tools, such as Hades, has no relation to packet loss experienced by user traffic (see Usage Examples on page 27). To identify what packet loss is experienced by user traffic, which may critically affect user satisfaction with network services, it must be monitored directly. It is not possible to extract any meaningful information from packet loss experienced by test packets Summary The Packetloss application has been developed and deployed on all -NREN links of partners participating in the SA3 pilot. This provides the following benefits: Real user traffic packet loss monitoring (rather than unrealistic packet loss that can be measured by test packets). Continuous, 100% non-intrusive monitoring. Plottable number of lost packets and packet loss rate for any user-specified time period with any time resolution. Document Code:

ABW - Short-timescale passive bandwidth monitoring

ABW - Short-timescale passive bandwidth monitoring ABW - Short-timescale passive bandwidth monitoring Sven Ubik (CESNET, Czech Republic), Demetres Antoniades (ICS-FORTH, Greece), Arne Oslebo (UNINETT, Norway) November 18, 2006 Abstract Bandwidth usage

More information

ABW - Short-timescale passive bandwidth monitoring

ABW - Short-timescale passive bandwidth monitoring ABW - Short-timescale passive bandwidth monitoring Sven Ubik (CESNET, Czech Republic), Demetres Antoniades (ICS-FORTH, Greece), Arne Oslebo (UNINETT, Norway) Abstract Bandwidth usage monitoring is important

More information

ABW Short-timescale passive bandwidth monitoring

ABW Short-timescale passive bandwidth monitoring CESNET technical report number 3/2007 ABW Short-timescale passive bandwidth monitoring Sven Ubik (CESNET), Demetres Antoniades (ICS-FORTH), Arne Oslebo (UNINETT) 7.12.2006 1 Abstract Bandwidth usage monitoring

More information

perfsonar MDM release 3.0 - Product Brief

perfsonar MDM release 3.0 - Product Brief perfsonar MDM release 3.0 - Product Brief In order to provide the fast, reliable and uninterrupted network communication that users of the GÉANT 2 research networks rely on, network administrators must

More information

Infrastructure for active and passive measurements at 10Gbps and beyond

Infrastructure for active and passive measurements at 10Gbps and beyond Infrastructure for active and passive measurements at 10Gbps and beyond Best Practice Document Produced by UNINETT led working group on network monitoring (UFS 142) Author: Arne Øslebø August 2014 1 TERENA

More information

HP IMC User Behavior Auditor

HP IMC User Behavior Auditor HP IMC User Behavior Auditor Administrator Guide Abstract This guide describes the User Behavior Auditor (UBA), an add-on service module of the HP Intelligent Management Center. UBA is designed for IMC

More information

Quick Start for Network Agent. 5-Step Quick Start. What is Network Agent?

Quick Start for Network Agent. 5-Step Quick Start. What is Network Agent? What is Network Agent? Websense Network Agent software monitors all internet traffic on the machines that you assign to it. Network Agent filters HTTP traffic and more than 70 other popular internet protocols,

More information

NetStream (Integrated) Technology White Paper HUAWEI TECHNOLOGIES CO., LTD. Issue 01. Date 2012-9-6

NetStream (Integrated) Technology White Paper HUAWEI TECHNOLOGIES CO., LTD. Issue 01. Date 2012-9-6 (Integrated) Technology White Paper Issue 01 Date 2012-9-6 HUAWEI TECHNOLOGIES CO., LTD. 2012. All rights reserved. No part of this document may be reproduced or transmitted in any form or by any means

More information

Realistic Passive Packet Loss Measurement for High-Speed Networks

Realistic Passive Packet Loss Measurement for High-Speed Networks Realistic Passive Packet Loss Measurement for High-Speed Networks Aleš Friedl 1, Sven Ubik 1, Alexandros Kapravelos 2, Michalis Polychronakis 2, and Evangelos P. Markatos 2 1 CESNET, Czech Republic {afriedl,ubik}@cesnet.cz

More information

SyncThru TM Web Admin Service Administrator Manual

SyncThru TM Web Admin Service Administrator Manual SyncThru TM Web Admin Service Administrator Manual 2007 Samsung Electronics Co., Ltd. All rights reserved. This administrator's guide is provided for information purposes only. All information included

More information

Comprehensive IP Traffic Monitoring with FTAS System

Comprehensive IP Traffic Monitoring with FTAS System Comprehensive IP Traffic Monitoring with FTAS System Tomáš Košňar kosnar@cesnet.cz CESNET, association of legal entities Prague, Czech Republic Abstract System FTAS is designed for large-scale continuous

More information

Experiences Deploying and Operating a Large-Scale Monitoring Infrastructure

Experiences Deploying and Operating a Large-Scale Monitoring Infrastructure 1 Experiences Deploying and Operating a Large-Scale Monitoring Infrastructure 25 th NORDUnet conference Arne Øslebø arne.oslebo@uninett.no Outline Background and motivation Typical setup Deployment map

More information

Network Probe User Guide

Network Probe User Guide Network Probe User Guide Network Probe User Guide Table of Contents 1. Introduction...1 2. Installation...2 Windows installation...2 Linux installation...3 Mac installation...4 License key...5 Deployment...5

More information

Visio Enabled Solution: One-Click Switched Network Vision

Visio Enabled Solution: One-Click Switched Network Vision Visio Enabled Solution: One-Click Switched Network Vision Tim Wittwer, Senior Software Engineer Alan Delwiche, Senior Software Engineer March 2001 Applies to: All Microsoft Visio 2002 Editions All Microsoft

More information

Figure 1. perfsonar architecture. 1 This work was supported by the EC IST-EMANICS Network of Excellence (#26854).

Figure 1. perfsonar architecture. 1 This work was supported by the EC IST-EMANICS Network of Excellence (#26854). 1 perfsonar tools evaluation 1 The goal of this PSNC activity was to evaluate perfsonar NetFlow tools for flow collection solution and assess its applicability to easily subscribe and request different

More information

Transport Layer Protocols

Transport Layer Protocols Transport Layer Protocols Version. Transport layer performs two main tasks for the application layer by using the network layer. It provides end to end communication between two applications, and implements

More information

D1.2 Network Load Balancing

D1.2 Network Load Balancing D1. Network Load Balancing Ronald van der Pol, Freek Dijkstra, Igor Idziejczak, and Mark Meijerink SARA Computing and Networking Services, Science Park 11, 9 XG Amsterdam, The Netherlands June ronald.vanderpol@sara.nl,freek.dijkstra@sara.nl,

More information

Using TrueSpeed VNF to Test TCP Throughput in a Call Center Environment

Using TrueSpeed VNF to Test TCP Throughput in a Call Center Environment Using TrueSpeed VNF to Test TCP Throughput in a Call Center Environment TrueSpeed VNF provides network operators and enterprise users with repeatable, standards-based testing to resolve complaints about

More information

Network Probe. Figure 1.1 Cacti Utilization Graph

Network Probe. Figure 1.1 Cacti Utilization Graph Network Probe Description The MCNC Client Network Engineering group will install several open source network performance management tools on a computer provided by the LEA or charter school to build a

More information

White Paper. The Ten Features Your Web Application Monitoring Software Must Have. Executive Summary

White Paper. The Ten Features Your Web Application Monitoring Software Must Have. Executive Summary White Paper The Ten Features Your Web Application Monitoring Software Must Have Executive Summary It s hard to find an important business application that doesn t have a web-based version available and

More information

BASIC ANALYSIS OF TCP/IP NETWORKS

BASIC ANALYSIS OF TCP/IP NETWORKS BASIC ANALYSIS OF TCP/IP NETWORKS INTRODUCTION Communication analysis provides powerful tool for maintenance, performance monitoring, attack detection, and problems fixing in computer networks. Today networks

More information

How To Create A Network Monitoring System (Flowmon) In Avea-Tech (For Free)

How To Create A Network Monitoring System (Flowmon) In Avea-Tech (For Free) Network Traffic Performance & Security Monitoring Project proposal minimal project Orsenna;Invea-Tech FLOWMON PROBES 1000 & 100 Contents 1. Introduction... 2 1.1. General System Requirements... 2 1.2.

More information

Deliverable D4.2a: First LOBSTER Workshop Proceedings and Summary

Deliverable D4.2a: First LOBSTER Workshop Proceedings and Summary INFORMATION SOCIETY TECHNOLOGIES (IST) PROGRAMME Large Scale Monitoring of BroadBand Internet Infrastructure Contract No. 004336 Deliverable D4.2a: First LOBSTER Workshop Proceedings and Summary Abstract:

More information

11.1. Performance Monitoring

11.1. Performance Monitoring 11.1. Performance Monitoring Windows Reliability and Performance Monitor combines the functionality of the following tools that were previously only available as stand alone: Performance Logs and Alerts

More information

Rebasoft Auditor Quick Start Guide

Rebasoft Auditor Quick Start Guide Copyright Rebasoft Limited: 2009-2011 1 Release 2.1, Rev. 1 Copyright Notice Copyright 2009-2011 Rebasoft Ltd. All rights reserved. REBASOFT Software, the Rebasoft logo, Rebasoft Auditor are registered

More information

SOLARWINDS ENGINEER S TOOLSET FAST FIXES TO NETWORK ISSUES

SOLARWINDS ENGINEER S TOOLSET FAST FIXES TO NETWORK ISSUES DATASHEET SOLARWINDS ENGINEER S TOOLSET FAST FIXES TO NETWORK ISSUES SolarWinds Engineer s Toolset (ETS) helps you monitor and troubleshoot your network with the most trusted tools in network management.

More information

Cisco Performance Visibility Manager 1.0.1

Cisco Performance Visibility Manager 1.0.1 Cisco Performance Visibility Manager 1.0.1 Cisco Performance Visibility Manager (PVM) is a proactive network- and applicationperformance monitoring, reporting, and troubleshooting system for maximizing

More information

SSL VPN Technology White Paper

SSL VPN Technology White Paper SSL VPN Technology White Paper Keywords: SSL VPN, HTTPS, Web access, TCP access, IP access Abstract: SSL VPN is an emerging VPN technology based on HTTPS. This document describes its implementation and

More information

FileNet System Manager Dashboard Help

FileNet System Manager Dashboard Help FileNet System Manager Dashboard Help Release 3.5.0 June 2005 FileNet is a registered trademark of FileNet Corporation. All other products and brand names are trademarks or registered trademarks of their

More information

Quareo ICM Server Software

Quareo ICM Server Software The Quareo Infrastructure Configuration Manager (ICM) is a server software application designed to document and administer both passive and active network connectivity infrastructure. ICM enables management

More information

Test Equipment Depot - 800.517.8431-99 Washington Street Melrose, MA 02176 - TestEquipmentDepot.com. Application Advisor

Test Equipment Depot - 800.517.8431-99 Washington Street Melrose, MA 02176 - TestEquipmentDepot.com. Application Advisor Test Equipment Depot - 800.517.8431-99 Washington Street Melrose, MA 02176 - TestEquipmentDepot.com NetAlly Application Advisor Monitor End User Experience for Local and Remote Users, Distributed Sites

More information

Hands on Workshop. Network Performance Monitoring and Multicast Routing. Yasuichi Kitamura NICT Jin Tanaka KDDI/NICT APAN-JP NOC

Hands on Workshop. Network Performance Monitoring and Multicast Routing. Yasuichi Kitamura NICT Jin Tanaka KDDI/NICT APAN-JP NOC Hands on Workshop Network Performance Monitoring and Multicast Routing Yasuichi Kitamura NICT Jin Tanaka KDDI/NICT APAN-JP NOC July 18th TEIN2 Site Coordination Workshop Network Performance Monitoring

More information

Network Management and Monitoring Software

Network Management and Monitoring Software Page 1 of 7 Network Management and Monitoring Software Many products on the market today provide analytical information to those who are responsible for the management of networked systems or what the

More information

Quick Start for Network Agent. 5-Step Quick Start. What is Network Agent?

Quick Start for Network Agent. 5-Step Quick Start. What is Network Agent? What is Network Agent? The Websense Network Agent software component uses sniffer technology to monitor all of the internet traffic on the network machines that you assign to it. Network Agent filters

More information

Instructions for Access to Summary Traffic Data by GÉANT Partners and other Organisations

Instructions for Access to Summary Traffic Data by GÉANT Partners and other Organisations Contract Number: IST-2000-26417 Project Title: Deliverable D8 : Instructions for Access to Summary Traffic Data by GÉANT Partners and other Organisations Contractual Date: 31 May 2002 Actual Date: 14 August

More information

RUNNING A HELPDESK CONTENTS. using HP Web Jetadmin

RUNNING A HELPDESK CONTENTS. using HP Web Jetadmin RUNNING A HELPDESK using HP Web Jetadmin CONTENTS Overview... 2 Helpdesk examples... 2 Viewing devices... 2 Quick Device Discovery... 3 Search... 3 Filters... 3 Columns... 4 Device Groups... 4 Troubleshooting

More information

Research on Errors of Utilized Bandwidth Measured by NetFlow

Research on Errors of Utilized Bandwidth Measured by NetFlow Research on s of Utilized Bandwidth Measured by NetFlow Haiting Zhu 1, Xiaoguo Zhang 1,2, Wei Ding 1 1 School of Computer Science and Engineering, Southeast University, Nanjing 211189, China 2 Electronic

More information

Question: 3 When using Application Intelligence, Server Time may be defined as.

Question: 3 When using Application Intelligence, Server Time may be defined as. 1 Network General - 1T6-521 Application Performance Analysis and Troubleshooting Question: 1 One component in an application turn is. A. Server response time B. Network process time C. Application response

More information

VMWARE WHITE PAPER 1

VMWARE WHITE PAPER 1 1 VMWARE WHITE PAPER Introduction This paper outlines the considerations that affect network throughput. The paper examines the applications deployed on top of a virtual infrastructure and discusses the

More information

perfsonar Multi-Domain Monitoring Service Deployment and Support: The LHC-OPN Use Case

perfsonar Multi-Domain Monitoring Service Deployment and Support: The LHC-OPN Use Case perfsonar Multi-Domain Monitoring Service Deployment and Support: The LHC-OPN Use Case Fausto Vetter, Domenico Vicinanza DANTE TNC 2010, Vilnius, 2 June 2010 Agenda Large Hadron Collider Optical Private

More information

TANDBERG MANAGEMENT SUITE 10.0

TANDBERG MANAGEMENT SUITE 10.0 TANDBERG MANAGEMENT SUITE 10.0 Installation Manual Getting Started D12786 Rev.16 This document is not to be reproduced in whole or in part without permission in writing from: Contents INTRODUCTION 3 REQUIREMENTS

More information

PANDORA FMS NETWORK DEVICE MONITORING

PANDORA FMS NETWORK DEVICE MONITORING NETWORK DEVICE MONITORING pag. 2 INTRODUCTION This document aims to explain how Pandora FMS is able to monitor all network devices available on the marke such as Routers, Switches, Modems, Access points,

More information

Limitations of Packet Measurement

Limitations of Packet Measurement Limitations of Packet Measurement Collect and process less information: Only collect packet headers, not payload Ignore single packets (aggregate) Ignore some packets (sampling) Make collection and processing

More information

HP Insight Management Agents architecture for Windows servers

HP Insight Management Agents architecture for Windows servers HP Insight Management Agents architecture for Windows servers Technology brief, 2 nd edition Introduction... 3 A first look at the Insight Management Agents architecture... 3 HP Insight Management agents...

More information

CREW - FP7 - GA No. 258301. Cognitive Radio Experimentation World. Project Deliverable D7.5.4 Showcase of experiment ready (Demonstrator)

CREW - FP7 - GA No. 258301. Cognitive Radio Experimentation World. Project Deliverable D7.5.4 Showcase of experiment ready (Demonstrator) Cognitive Radio Experimentation World!"#$% Project Deliverable Showcase of experiment ready (Demonstrator) Contractual date of delivery: 31-03-14 Actual date of delivery: 18-04-14 Beneficiaries: Lead beneficiary:

More information

Cisco IOS Flexible NetFlow Technology

Cisco IOS Flexible NetFlow Technology Cisco IOS Flexible NetFlow Technology Last Updated: December 2008 The Challenge: The ability to characterize IP traffic and understand the origin, the traffic destination, the time of day, the application

More information

CCNA Discovery 4.0.3.0 Networking for Homes and Small Businesses Student Packet Tracer Lab Manual

CCNA Discovery 4.0.3.0 Networking for Homes and Small Businesses Student Packet Tracer Lab Manual 4.0.3.0 Networking for Homes and Small Businesses Student Packet Tracer Lab Manual This document is exclusive property of Cisco Systems, Inc. Permission is granted to print and copy this document for non-commercial

More information

Using Fuzzy Logic Control to Provide Intelligent Traffic Management Service for High-Speed Networks ABSTRACT:

Using Fuzzy Logic Control to Provide Intelligent Traffic Management Service for High-Speed Networks ABSTRACT: Using Fuzzy Logic Control to Provide Intelligent Traffic Management Service for High-Speed Networks ABSTRACT: In view of the fast-growing Internet traffic, this paper propose a distributed traffic management

More information

Tue Apr 19 11:03:19 PDT 2005 by Andrew Gristina thanks to Luca Deri and the ntop team

Tue Apr 19 11:03:19 PDT 2005 by Andrew Gristina thanks to Luca Deri and the ntop team Tue Apr 19 11:03:19 PDT 2005 by Andrew Gristina thanks to Luca Deri and the ntop team This document specifically addresses a subset of interesting netflow export situations to an ntop netflow collector

More information

Kiwi SyslogGen. A Freeware Syslog message generator for Windows. by SolarWinds, Inc.

Kiwi SyslogGen. A Freeware Syslog message generator for Windows. by SolarWinds, Inc. Kiwi SyslogGen A Freeware Syslog message generator for Windows by SolarWinds, Inc. Kiwi SyslogGen is a free Windows Syslog message generator which sends Unix type Syslog messages to any PC or Unix Syslog

More information

Applications. Network Application Performance Analysis. Laboratory. Objective. Overview

Applications. Network Application Performance Analysis. Laboratory. Objective. Overview Laboratory 12 Applications Network Application Performance Analysis Objective The objective of this lab is to analyze the performance of an Internet application protocol and its relation to the underlying

More information

LOCKSS on LINUX. CentOS6 Installation Manual 08/22/2013

LOCKSS on LINUX. CentOS6 Installation Manual 08/22/2013 LOCKSS on LINUX CentOS6 Installation Manual 08/22/2013 1 Table of Contents Overview... 3 LOCKSS Hardware... 5 Installation Checklist... 6 BIOS Settings... 9 Installation... 10 Firewall Configuration...

More information

J-Flow on J Series Services Routers and Branch SRX Series Services Gateways

J-Flow on J Series Services Routers and Branch SRX Series Services Gateways APPLICATION NOTE Juniper Flow Monitoring J-Flow on J Series Services Routers and Branch SRX Series Services Gateways Copyright 2011, Juniper Networks, Inc. 1 APPLICATION NOTE - Juniper Flow Monitoring

More information

Cisco Application Networking Manager Version 2.0

Cisco Application Networking Manager Version 2.0 Cisco Application Networking Manager Version 2.0 Cisco Application Networking Manager (ANM) software enables centralized configuration, operations, and monitoring of Cisco data center networking equipment

More information

Introduction to perfsonar

Introduction to perfsonar Introduction to perfsonar Loukik Kudarimoti, DANTE 27 th September, 2006 SEEREN2 Summer School, Heraklion Overview of this talk Answers to some basic questions The need for Multi-domain monitoring What

More information

ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy

ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy ZEN LOAD BALANCER EE v3.04 DATASHEET The Load Balancing made easy OVERVIEW The global communication and the continuous growth of services provided through the Internet or local infrastructure require to

More information

NetCrunch 6. AdRem. Network Monitoring Server. Document. Monitor. Manage

NetCrunch 6. AdRem. Network Monitoring Server. Document. Monitor. Manage AdRem NetCrunch 6 Network Monitoring Server With NetCrunch, you always know exactly what is happening with your critical applications, servers, and devices. Document Explore physical and logical network

More information

VisuSniff: A Tool For The Visualization Of Network Traffic

VisuSniff: A Tool For The Visualization Of Network Traffic VisuSniff: A Tool For The Visualization Of Network Traffic Rainer Oechsle University of Applied Sciences, Trier Postbox 1826 D-54208 Trier +49/651/8103-508 oechsle@informatik.fh-trier.de Oliver Gronz University

More information

REQUIREMENTS AND INSTALLATION OF THE NEFSIS DEDICATED SERVER

REQUIREMENTS AND INSTALLATION OF THE NEFSIS DEDICATED SERVER NEFSIS TRAINING SERIES Nefsis Dedicated Server version 5.1.0.XXX Requirements and Implementation Guide (Rev 4-10209) REQUIREMENTS AND INSTALLATION OF THE NEFSIS DEDICATED SERVER Nefsis Training Series

More information

SAIP 2012 Performance Engineering

SAIP 2012 Performance Engineering SAIP 2012 Performance Engineering Author: Jens Edlef Møller (jem@cs.au.dk) Instructions for installation, setup and use of tools. Introduction For the project assignment a number of tools will be used.

More information

Enabling NAT and Routing in DGW v2.0 June 6, 2012

Enabling NAT and Routing in DGW v2.0 June 6, 2012 Enabling NAT and Routing in DGW v2.0 June 6, 2012 Proprietary 2012 Media5 Corporation Table of Contents Introduction... 3 Starting Services... 4 Distinguishing your WAN and LAN interfaces... 5 Configuring

More information

GN1 (GÉANT) Deliverable D13.2

GN1 (GÉANT) Deliverable D13.2 Contract Number: IST-2000-26417 Project Title: GN1 (GÉANT) Deliverable 13.2 Technology Roadmap for Year 3 Contractual Date: 31 July 2002 Actual Date: 16 August 2002 Work Package: WP8 Nature of Deliverable:

More information

Hillstone StoneOS User Manual Hillstone Unified Intelligence Firewall Installation Manual

Hillstone StoneOS User Manual Hillstone Unified Intelligence Firewall Installation Manual Hillstone StoneOS User Manual Hillstone Unified Intelligence Firewall Installation Manual www.hillstonenet.com Preface Conventions Content This document follows the conventions below: CLI Tip: provides

More information

Features Overview Guide About new features in WhatsUp Gold v12

Features Overview Guide About new features in WhatsUp Gold v12 Features Overview Guide About new features in WhatsUp Gold v12 Contents CHAPTER 1 Learning about new features in Ipswitch WhatsUp Gold v12 Welcome to WhatsUp Gold... 1 What's new in WhatsUp Gold v12...

More information

Step-by-Step Configuration

Step-by-Step Configuration Step-by-Step Configuration Kerio Technologies C 2001-2003 Kerio Technologies. All Rights Reserved. Printing Date: December 17, 2003 This guide provides detailed description on configuration of the local

More information

Implementing Network Monitoring Tools

Implementing Network Monitoring Tools Section 1 Network Systems Engineering Implementing Network Monitoring Tools V.C.Asiwe and P.S.Dowland Network Research Group, University of Plymouth, Plymouth, United Kingdom e-mail: info@network-research-group.org

More information

Lab 1: Packet Sniffing and Wireshark

Lab 1: Packet Sniffing and Wireshark Introduction CSC 5991 Cyber Security Practice Lab 1: Packet Sniffing and Wireshark The first part of the lab introduces packet sniffer, Wireshark. Wireshark is a free opensource network protocol analyzer.

More information

Lab - Dual Boot - Vista & Windows XP

Lab - Dual Boot - Vista & Windows XP Lab - Dual Boot - Vista & Windows XP Brought to you by RMRoberts.com After completing this lab activity, you will be able to: Install and configure a dual boot Windows XP and Vista operating systems. Explain

More information

Chapter 3. Internet Applications and Network Programming

Chapter 3. Internet Applications and Network Programming Chapter 3 Internet Applications and Network Programming 1 Introduction The Internet offers users a rich diversity of services none of the services is part of the underlying communication infrastructure

More information

PANDORA FMS NETWORK DEVICES MONITORING

PANDORA FMS NETWORK DEVICES MONITORING NETWORK DEVICES MONITORING pag. 2 INTRODUCTION This document aims to explain how Pandora FMS can monitor all the network devices available in the market, like Routers, Switches, Modems, Access points,

More information

McAfee Web Gateway 7.4.1

McAfee Web Gateway 7.4.1 Release Notes Revision B McAfee Web Gateway 7.4.1 Contents About this release New features and enhancements Resolved issues Installation instructions Known issues Find product documentation About this

More information

Cisco NetFlow Reporting Instruction Manual Version 1.0

Cisco NetFlow Reporting Instruction Manual Version 1.0 Cisco NetFlow Reporting Instruction Manual Version 1.0 WiredCity 777 Davis St, Suite 250 San Leandro CA 94577 Ph: + 1 510 297 5874 Fax: +1 510-357-8136 itmonitor@wiredcity.com www.wiredcity.com www.wiredcity.com

More information

How A V3 Appliance Employs Superior VDI Architecture to Reduce Latency and Increase Performance

How A V3 Appliance Employs Superior VDI Architecture to Reduce Latency and Increase Performance How A V3 Appliance Employs Superior VDI Architecture to Reduce Latency and Increase Performance www. ipro-com.com/i t Contents Overview...3 Introduction...3 Understanding Latency...3 Network Latency...3

More information

Chapter 6 Virtual Private Networking Using SSL Connections

Chapter 6 Virtual Private Networking Using SSL Connections Chapter 6 Virtual Private Networking Using SSL Connections The FVS336G ProSafe Dual WAN Gigabit Firewall with SSL & IPsec VPN provides a hardwarebased SSL VPN solution designed specifically to provide

More information

Edge Configuration Series Reporting Overview

Edge Configuration Series Reporting Overview Reporting Edge Configuration Series Reporting Overview The Reporting portion of the Edge appliance provides a number of enhanced network monitoring and reporting capabilities. WAN Reporting Provides detailed

More information

High-Performance IP Service Node with Layer 4 to 7 Packet Processing Features

High-Performance IP Service Node with Layer 4 to 7 Packet Processing Features UDC 621.395.31:681.3 High-Performance IP Service Node with Layer 4 to 7 Packet Processing Features VTsuneo Katsuyama VAkira Hakata VMasafumi Katoh VAkira Takeyama (Manuscript received February 27, 2001)

More information

The OSI model has seven layers. The principles that were applied to arrive at the seven layers can be briefly summarized as follows:

The OSI model has seven layers. The principles that were applied to arrive at the seven layers can be briefly summarized as follows: 1.4 Reference Models Now that we have discussed layered networks in the abstract, it is time to look at some examples. In the next two sections we will discuss two important network architectures, the

More information

EKT 332/4 COMPUTER NETWORK

EKT 332/4 COMPUTER NETWORK UNIVERSITI MALAYSIA PERLIS SCHOOL OF COMPUTER & COMMUNICATIONS ENGINEERING EKT 332/4 COMPUTER NETWORK LABORATORY MODULE LAB 2 NETWORK PROTOCOL ANALYZER (SNIFFING AND IDENTIFY PROTOCOL USED IN LIVE NETWORK)

More information

Lab - Observing DNS Resolution

Lab - Observing DNS Resolution Objectives Part 1: Observe the DNS Conversion of a URL to an IP Address Part 2: Observe DNS Lookup Using the nslookup Command on a Web Site Part 3: Observe DNS Lookup Using the nslookup Command on Mail

More information

Configuring a SIP Trunk between Avaya Aura Session Manager Release 6.1 and Avaya Communication Server 1000E Release 7.5 Issue 1.0

Configuring a SIP Trunk between Avaya Aura Session Manager Release 6.1 and Avaya Communication Server 1000E Release 7.5 Issue 1.0 Avaya Solution Interoperability Test Lab Configuring a SIP Trunk between Avaya Aura Session Manager Release 6.1 and Avaya Communication Server 1000E Release 7.5 Issue 1.0 Abstract These Application Notes

More information

CHAPTER. Monitoring and Diagnosing

CHAPTER. Monitoring and Diagnosing CHAPTER 20. This chapter provides details about using the Diagnostics & Monitoring system available through ShoreTel Director. It contains the following information: Overview... 661 Architecture... 661

More information

Router Architectures

Router Architectures Router Architectures An overview of router architectures. Introduction What is a Packet Switch? Basic Architectural Components Some Example Packet Switches The Evolution of IP Routers 2 1 Router Components

More information

Open Source in Network Administration: the ntop Project

Open Source in Network Administration: the ntop Project Open Source in Network Administration: the ntop Project Luca Deri 1 Project History Started in 1997 as monitoring application for the Univ. of Pisa 1998: First public release v 0.4 (GPL2) 1999-2002:

More information

Barracuda Link Balancer

Barracuda Link Balancer Barracuda Networks Technical Documentation Barracuda Link Balancer Administrator s Guide Version 2.2 RECLAIM YOUR NETWORK Copyright Notice Copyright 2004-2011, Barracuda Networks www.barracuda.com v2.2-110503-01-0503

More information

Cisco NetFlow TM Briefing Paper. Release 2.2 Monday, 02 August 2004

Cisco NetFlow TM Briefing Paper. Release 2.2 Monday, 02 August 2004 Cisco NetFlow TM Briefing Paper Release 2.2 Monday, 02 August 2004 Contents EXECUTIVE SUMMARY...3 THE PROBLEM...3 THE TRADITIONAL SOLUTIONS...4 COMPARISON WITH OTHER TECHNIQUES...6 CISCO NETFLOW OVERVIEW...7

More information

Dell SupportAssist Version 2.0 for Dell OpenManage Essentials Quick Start Guide

Dell SupportAssist Version 2.0 for Dell OpenManage Essentials Quick Start Guide Dell SupportAssist Version 2.0 for Dell OpenManage Essentials Quick Start Guide Notes, Cautions, and Warnings NOTE: A NOTE indicates important information that helps you make better use of your computer.

More information

Autonomous NetFlow Probe

Autonomous NetFlow Probe Autonomous Ladislav Lhotka lhotka@cesnet.cz Martin Žádník xzadni00@stud.fit.vutbr.cz TF-CSIRT meeting, September 15, 2005 Outline 1 2 Specification Hardware Firmware Software 3 4 Short-term fixes Test

More information

DIGICLIENT 8.0 Remote Agent Software

DIGICLIENT 8.0 Remote Agent Software DIGICLIENT 8.0 Remote Agent Software MODEL: D17800 Series Instruction Manual English Version 1.0 Copyright 2007 Digimerge Technologies Inc Table of Contents Table of Contents About the DigiClient 8.0...

More information

G DATA TechPaper #0275. G DATA Network Monitoring

G DATA TechPaper #0275. G DATA Network Monitoring G DATA TechPaper #0275 G DATA Network Monitoring G DATA Software AG Application Development May 2016 Contents Introduction... 3 1. The benefits of network monitoring... 3 1.1. Availability... 3 1.2. Migration

More information

Measuring IP Performance. Geoff Huston Telstra

Measuring IP Performance. Geoff Huston Telstra Measuring IP Performance Geoff Huston Telstra What are you trying to measure? User experience Responsiveness Sustained Throughput Application performance quality Consistency Availability Network Behaviour

More information

COMMANDS 1 Overview... 1 Default Commands... 2 Creating a Script from a Command... 10 Document Revision History... 10

COMMANDS 1 Overview... 1 Default Commands... 2 Creating a Script from a Command... 10 Document Revision History... 10 LabTech Commands COMMANDS 1 Overview... 1 Default Commands... 2 Creating a Script from a Command... 10 Document Revision History... 10 Overview Commands in the LabTech Control Center send specific instructions

More information

Cover. White Paper. (nchronos 4.1)

Cover. White Paper. (nchronos 4.1) Cover White Paper (nchronos 4.1) Copyright Copyright 2013 Colasoft LLC. All rights reserved. Information in this document is subject to change without notice. No part of this document may be reproduced

More information

SiteCelerate white paper

SiteCelerate white paper SiteCelerate white paper Arahe Solutions SITECELERATE OVERVIEW As enterprises increases their investment in Web applications, Portal and websites and as usage of these applications increase, performance

More information

COMPARATIVE EVALUATION OF AVAILABLE BANDWIDTH ESTIMATION TOOLS (PRTG AND OAUNETMON) IN A CAMPUS WIDE AREA NETWORK.

COMPARATIVE EVALUATION OF AVAILABLE BANDWIDTH ESTIMATION TOOLS (PRTG AND OAUNETMON) IN A CAMPUS WIDE AREA NETWORK. COMPARATIVE EVALUATION OF AVAILABLE BANDWIDTH ESTIMATION TOOLS (PRTG AND OAUNETMON) IN A CAMPUS WIDE AREA NETWORK. Azeez Nureni Ayofe Department of Maths & Computer Science, College of Natural and Applied

More information

mbits Network Operations Centrec

mbits Network Operations Centrec mbits Network Operations Centrec The mbits Network Operations Centre (NOC) is co-located and fully operationally integrated with the mbits Service Desk. The NOC is staffed by fulltime mbits employees,

More information

138 Configuration Wizards

138 Configuration Wizards 9 Configuration Wizards 9.1 Introduction to Wizards ACP ThinManager uses wizards for configuration. Wizards take two forms. List Wizards associate Terminal Servers and ThinManager Servers with their IP

More information

Chapter 8 Monitoring and Logging

Chapter 8 Monitoring and Logging Chapter 8 Monitoring and Logging This chapter describes the SSL VPN Concentrator status information, logging, alerting and reporting features. It describes: SSL VPN Concentrator Status Active Users Event

More information

www.novell.com/documentation Jobs Guide Identity Manager 4.0.1 February 10, 2012

www.novell.com/documentation Jobs Guide Identity Manager 4.0.1 February 10, 2012 www.novell.com/documentation Jobs Guide Identity Manager 4.0.1 February 10, 2012 Legal Notices Novell, Inc. makes no representations or warranties with respect to the contents or use of this documentation,

More information

General system requirements

General system requirements 2 General system requirements Minimal requirements Processor: Intel Core 2 Duo or equivalent Memory (RAM): HDD: NIC: 1 GB At least 100 MB available hard disk space. 1000 Mb/s, Jumbo frame 9kb. OS: Windows

More information

Configuring Your Gateman Proxy Server

Configuring Your Gateman Proxy Server Configuring Your Gateman Proxy Server A proxy server acts as an intermediary between a workstation users and the Internet to ensure security, administrative control, distribution of bandwidth and caching

More information