D6.7 Demonstration of international BoD connectivity at 4 Gb/s

Size: px
Start display at page:

Download "D6.7 Demonstration of international BoD connectivity at 4 Gb/s"

Transcription

1 D6.7 Demonstration of international BoD connectivity at 4 Gb/s Deliverable Number 07 WP # 6 Date 29th March 2012 Authors Jimmy Cullen, The University of Manchester Ralph Spencer, The University of Manchester Paul Boven, JIVE Fedde Bloemhof, JIVE Richard Hughes-Jones, DANTE Neal Jackson, The University of Manchester Paul Burgess, The University of Manchester Document Log Version Date Summary Fixed figure numbering, consistency of abbreviations, formatting Integrated NSI test results Final draft from Manchester Document created 1

2 Table of Contents D6.7 Demonstration of international BoD connectivity at 4 Gb/s Introduction Technologies AutoBAHN BoD Circuits NSI BoD Circuit Test Software Test Routes BoD Reservation Interface Tests Performed Results Conclusion Acknowledgements References

3 1 Introduction Very Long Baseline Interferometry (VLBI) is a technique in radio astronomy where radio telescopes at large geographical distances from one another carry out a coordinated observation, creating a virtual radio telescope of truly global scale and unsurpassed resolution. A typical observation lasts for 8 hours and nowadays generates about 4TB of data at each of the participating telescopes. The data from all these telescopes is then shipped to a central processing site where a special purpose supercomputer ('correlator') processes the data. Originally the data was recorded on magnetic tape and shipped, and more recently VLBI has started using standardized packs of hard-disks to ship this data between telescopes and correlator. The EXPReS project pioneered the use of high performance networks to transmit the data in real time, from telescopes around the globe to the correlator at JIVE in the Netherlands. This method of real-time observing is called e-vlbi. Given the high data rates involved, this generally requires dedicated networking connections between telescopes and correlator center. This traffic is mostly carried by the National Research and Education Networks (NRENs), who have built very advanced high speed networks. Provisioning these international paths across administrative boundaries requires a lot of coordination, hence the network for evlbi currently consists of static paths called 'light-paths', even though telescopes are not performing VLBI observations all the time. Bandwidth on Demand (BoD) is a service that makes it possible to reserve and provision dedicated paths between two points on a network, with a fixed bandwidth and duration. By dynamically creating and tearing down links for varying endpoints, many different users can share the networking resources, thus increasing the usage of the links. The complexity of the BoD system increases when endpoints lie within different NRENs, necessitating the interoperation of separate systems. BoD on international scale would be a very good match to the network usage patterns of e-vlbi, allowing for a more efficient and flexible use of scarce international networking resources. BoD on an international scale is a novel field, and not available yet as a production service. The testbeds that are currently available only offer limited bandwidth and are meant as a proof-of-concept. But the value of BoD services for end-users is specifically the very high bandwidth paths they offer. For (e-)vlbi, the sensitivity increases with the available bandwidth, and the European VLBI Network is currently growing from 1Gb/s towards generating 4Gb/s per telescope. Upgrading these BoD-testbeds and the connectivity of the participating institutes into these testbeds to support these speeds was a major undertaking. We encountered several unforeseen delays, both in the development of the standards and software, and in building out the network, which unfortunately led to this workpackage being delayed from its planned completion. This deliverable demonstrates that it is now possible to provision dedicated networking paths of 4Gb/s between several locations within Europe. Performing BoD tests at these speeds has helped push the state of the art forward and demonstrates that BoD may become a valuable tool in the e-sciences. 3

4 2 Technologies Several NRENs offer their end-users a BoD service. As different NRENs generally use equipment of different vendors and even have different underlying transport technologies for their network, these BoD services only span a single administrative domain. Examples of such BoD services would be OSCARs and OpenDRAC. When end-users, located in two separate NRENs, require connectivity, one of the major problems to address is the coordination of the allocation and reconfiguration of network resources across multiple administrative domains. The GÉANT network is an international network connecting many European NRENs. The BoD system offered by GÉANT is the pilot system Automated Bandwidth Allocation across Heterogeneous Networks (AutoBAHN) [1], which provides inter NREN BoD connectivity using the Inter-Domain Controller (IDC) protocol. Reservations are made by users from the AutoBAHN Client Portal [2], a web service provided on the GÉANT website. AutoBAHN provides an integrated business layer for coordinated inter-domain provisioning rather than providing a replacement for existing services. Since the NREN networks are heterogeneous, AutoBAHN is designed to match the capabilities of the interfacing networks for the provisioning of Layer 1 and 2 services. The Open Grid Forum, in collaboration with NRENs, end-users and industry, are currently working to create a standardised protocol for the delivery of BoD services called the Network Services Infrastructure (NSI) [3,4]. In the NEXPReS project it has been decided that this emerging BoD technology, which is envisaged to become the international standard, and therefore supported globally, will be the BoD technology used in transferring e-vlbi data. Many BoD services already offer an NSI interface (e.g the SURFnet/Netherlight OpenDRAC, the OpenNSA at NORDUnet/Northernlight) and others are implementing or planning to implement NSI as the standard matures (e.g. AutoBAHN and OSCARs). We realized early on that the nascent state of international BoD presented a risk for the timely completion of this deliverable, and it was decided to use a two-pronged approach: to try and use both AutoBAHN and NSI to achieve BoD at 4Gb/s for this deliverable. As most NRENs offer only one of the two BoD interfaces at this time, this would not cause a competition for resources, and would increase our chances of completing the deliverable. It turned out that both systems became able to support 4Gb/s traffic at almost exactly the same point in time, and we are therefore presenting the results from both tracks side by side. 4

5 3 AutoBAHN BoD Circuits In this demonstration, we used six end hosts located at four locations across three countries. The end hosts were located in three separate NRENs: jbnexpres1 located in the Schuster Building, University of Manchester, UK (JANET) jbnexpres2 located at the Jodrell Bank Observatory, University of Manchester, UK (JANET) EXPReS1 located in the GÉANT Point of Presence (PoP), London, UK EXPReS2 located in the Stockholm PoP, Sweden (NORDUnet) ara located in Metsähovi Radio Observatory, Aalto University, Finland (Funet) watt located in Metsähovi Radio Observatory, Aalto University, Finland (Funet) The PCs EXPReS1 and EPXReS2 were used in the EXPReS project for link testing, and were re-used in this project. Although the machine names refer to the previous project, all tests reported on in this document were done as part of NEXPReS. Figure 1 shows the network connecting hosts on the AutoBAHN system for this demonstration. Figure 1. Network map of NREN and end host connections to the GÉANT BoD core dynamic network. The blue routers are part of the BoD dynamic core and the grey routers part of the GÉANT static network. The clouds represent NREN resources, and computers represent end host locations. This network configuration is a mixture of static light paths and dynamic BoD paths, since BoD infrastructure and services are not provided yet on the whole path between the end hosts. 5

6 JANET, the UK's education and research network, provided a static light path from The University of Manchester to the GÉANT PoP in London at 4 Gb/s. In Manchester, the university has its own 10 Gb/s static network that allows end hosts to be located in the Schuster Building (School of Physics and Astronomy) at the main university campus and the Jodrell Bank Observatory, located 30 km south of the main campus. In the GÉANT PoP in London, an end host was located which connected directly to the AutoBAHN network. The GÉANT PoP in London connected to the PoP in Amsterdam over the dynamic core, then on to the Copenhagen PoP over a static path. In Copenhagen NORDUnet provided two 4 Gb/s static links, one to the Stockholm PoP and another to the administrative border with Funet. Funet then provided a static 4 Gb/s link through to the Metsähovi Radio Observatory. 3.1 Virtual Local Area Network (VLAN) Tagging The AutoBAHN system works via creating virtual LANs (VLANs) for data-streams traversing the network. The data is tagged with a VLAN tag (formally described in IEEE 802.1Q [5]), an optional 32 bit field that is added to the Ethernet frame and used by VLAN aware equipment to identify which VLAN a packet belongs to. AutoBAHN was configured to use several VLAN tags to identify sections of the links between end hosts. For example the connection between EXPReS1 and the BoD network was assigned the VLAN tag 2001, and therefore only Ethernet frames containing this tag would be delivered to EXPReS1. The AutoBAHN system dynamically alters VLAN tags based upon the BoD paths requested by the user. For example, if a path was requested from Metsähovi to Jodrell Bank, Ethernet frames sent over the Funet network are tagged with the ID When these arrive at the BoD network, via NORDUnet, AutoBAHN dynamically changes this tag value to 2002, which is the identifier for the VLAN across JANET to Jodrell Bank. 6

7 4 NSI BoD Circuit For our tests with NSI, we provisioned a circuit from Onsala Space Observatory (OSO) in Sweden, to JIVE in the Netherlands. Real-time transfers of 4Gb/s from OSO to JIVE will closely resemble the real traffic that we intend to use BoD for. During the work for this workpackage, the path was built out and tested in stages. We made use of the following end-hosts: JIVE (various hosts) located in Dwingeloo, the Netherlands iperf1.amsterdam1 located in Amsterdam, the Netherlands (SURFnet/Netherlight) EXPReS2 located in the Stockholm PoP, Sweden (NORDUnet) NEXPReS located in Onsala Space Observatory (OSO), Sweden (NORDUnet/SUNET) We made use of two NSI instances to direct the traffic: 1. OpenDRAC running on the SURFnet/Netherlight equipment in Amsterdam 2. OpenNSA running on the NORDUnet/Northernlight equipment in Stockholm SUNET configured an extra VLAN from OSO into NORDUnet, and SURFnet provided an extra 10G link from Dwingeloo to Amsterdam, so that the end-point hosts could connect to the dynamic part of the network. The connection from OSO to NORDUnet re-used the 10G interface that was in use for EXPReS2, causing that machine to be disconnected. The contention for this interface caused some unexpected delay in completing the work, as the NSI tests couldn't be performed before the AutoBAHN tests were finished. We used the NEXPReS NSI client prototype that was delivered in D6.01 of this workpackage to create the reservations for the paths. Our NSI client connected to both the SURFnet OpenDRAC, and the NORDUnet OpenNSA, to request each leg of the path from the NSA responsible for that administrative domain. Figure 2: Networking equipment, links and hosts used for the NSI BoD tests. 7

8 5 Test Software e-vlbi data is transferred using the User Datagram Protocol (UDP) since this is most appropriate for the high bandwidth constant bit rate data generated at radio telescopes [6,7,8,9]. UDPmon is software that was written specifically to test high speed data links with UDP data, and uses pseudo-random data to simulate real VLBI data. UDPmon [10] is a command line tool written for Linux operating systems. A website has been developed at The University of Manchester for performing network bandwidth tests between pre-configured end hosts using the UDPmon software. This work is part of Task 3 of NEXPReS WP6. Figure 3 shows two screen shots of the website. Figure 3. Two screenshots of the network test website. The top image shows the basic test page, where the user selects the client (sending) and server (receiving) machines, the size and number of packets to send, the ports to use on the end hosts and the socket buffer size. The lower image shows the results of a graphical network test, where received wire rate is plotted against inter-packet spacing for a range of values. For the tests using the AutoBAHN system, the website was used to verify connectivity and bandwidth between end hosts. For more detailed tests however, the native software was run from the command line as this allows all aspects of the software to be employed. 8

9 6 Test Routes The tests performed over the GÉANT AutoBAHN network connected four destinations in three countries: 1. Jodrell Bank Observatory/University of Manchester, UK, 2. GÉANT PoP, London, UK, 3. Stockholm PoP, Sweden, and 4. Metsähovi Radio Observatory, Aalto University, Finland At The University of Manchester there are two end hosts, jbnexpres1 situated in the Schuster Building at the main campus, and jbnexpres2 situated at Jodrell Bank Observatory. At Metsähovi Radio Observatory there are also two end hosts, watt and ara, both located at the observatory. Figure 4 is a screen shot from the network test website, showing a map of the locations involved in the BoD network tests. Figure 4. Map from the network test website, showing the location of the end hosts on the AutoBAHN and NSI tests. Sites involved in the AutoBAHN BoD are represented by the dark blue (University of Manchester), bright green (GÉANT PoPs), and bright red (Aalto University) markers. The orange (JIVE), light blue (Onsala) and right hand bright green (GÉANT Stockholm PoP) markers identify those sites involved in the NSI BoD test. The network interface used by the host EXPReS2 in the Stockholm PoP was used for both the AutoBAHN and NSI client test. To switch between these different technologies required changes to the physical interfaces which meant that only one BoD architecture could be used at a time. 9

10 7 BoD Reservation Interface Both AutoBAHN and NSI offer the end-user the ability to request a network path between two points on their network. For AutoBAHN, the user interface used in these tests was their client web-site. For NSI, the reservations were made by sending the reservations out using the NSI protocol, using the NEXPReS NSI implementation developed in this workpackage. 7.1 AutoBAHN BoD Reservation Interface Figure 4 shows a screen shot of the AutoBAHN Client Portal, used to make BoD reservations on the network. Reservations are created through the web service available to approved users on the GÉANT website. The reservations created here are translated in real time to changes in the topology of the BoD core network. Figure 5 Screen shot of the AutoBAHN Client Portal, a web service used to dynamically alter the configuration of the BoD network. 10

11 7.2 NSI BoD reservation Interface The NSI protocol, our implementation of an NSI client, and the user interface we built to let end-users interact with NSI, are all documented in our report for deliverable D6.01 [11]. Figure 6 shows a screenshot of the user interface, captured while we were testing the link from OSO to JIVE at 4Gb/s. It shows that we have both a reservation within the Netherlight domain, and one in the Northernlight domain. These two links together create the end-to-end path. Figure 6: NSI client user-interface, showing provisioned paths during one of the tests. 11

12 8 Tests Performed Detailed analysis of the network characteristics inevitably requires detailed measurements, which can take several hours to perform. This timescale, while not a limitation itself, does not easily lend itself to web-based interfaces, where the user usually expects a timely response. Therefore the network test website was used for simple validation of the connectivity of the BoD network, available bandwidth and any packet losses, but further detailed tests were run from the command line on the Linux end hosts. Received wire rate, which measures payload (user data) plus network protocol metadata (headers), more accurately describes the amount of data that the network is handling than received data rate, which only measures the payload. Packet loss is also a very important metric of the network characterization, which informs upon the fidelity of the link. Running tests at various packet sizes is another important network performance measure as, for a constant data rate, the packet size is inversely proportional to the number of packets handled by the network. Packet jitter is described as the variation in inter-packet arrival times from a constant data rate source. It is important that we know if and how the packets are delayed or accelerated so that phenomena such as packet reordering and bursting can be identified. In order to investigate these characteristics tests were performed between end hosts where a set number of packets were sent and received. For a specified packet size, the inter-packet spacing was varied between well-defined values at small intervals to allow investigation of received wire rate and packet loss. Tests were repeated for various packet sizes. In the tests, for a given packet size, 10,000 packets were sent with requested inter-packet spacing between 0 and 40 microseconds at 0.1 microsecond intervals. Ten different packet sizes were used, ranging from 1000 to 8972 bytes, which was the maximum value for the line. Therefore for each set of tests between end hosts, 40 million packets were transmitted and analysed. These network tests were performed between: (a) (b) (c) (d) (e) (f) London (EXPReS1) and Stockholm (EXPReS2) Stockholm (EXPReS2) and London (EXPReS1) Stockholm (EXPReS2) and Jodrell Bank (jbnexpres2) Jodrell Bank (jbnexpres2) and Stockholm (EXPReS2) Metsahovi (ara) and Jodrell Bank (jbnexpres2) Jodrell Bank (jbnexpres2) and Metsahovi (watt) Finally, to simulate a real e-vlbi experiment, two 24 hour transfers of data were performed between jbnexpres2 and EXPReS2. Bandwidth, packets lost and packets reordered were measured. 12

13 9 Results In this section we present the results of the network tests for the both BoD architectures. Sections give the AutoBAHN BoD circuit results and 9.6 gives the results of the NSI BoD tests. 9.1 Received Wire Rate Figure 7 plots inter-packet spacing versus received wire rate for each of the ten packet size values, for each of the 6 links tested. The plots show the requested inter-packet spacing on the abscissa with the received wire rate on ordinate. Different packet sizes are denoted by different format and coloured lines. (a) (b) (c) (d) (e) (f) Figure 7: Requested inter-packet spacing versus received wire rate for a range of UDP packet sizes for each of the six links tested. 13

14 9.2 Percentage Packet Loss Figure 8 plots inter-packet spacing versus the percentage packet loss results for each packet size tested over the six test routes. Again, different packet sizes are denoted by different format and coloured lines. (a) (b) (c) (d) (e) (f) Figure 8: Requested inter-packet spacing versus percentage packet loss for each of the ten packet sizes tested, over each of the six routes. Figure sub-elements correspond to those in figure 7. 14

15 9.3 Jitter Here we present the inter-packet jitter for various links at three inter-packet spacing, representing 1, 2 and 4 Gb/s throughput. In each test 106 packets were sent and recorded. (a) (b) (c) (d) Figure 9: Histograms displaying the packet jitter for one million packets sent at 1, 2 and 4 Gb/s. The panel show the routes (a) London to Stockholm, (b) Stockholm to London, (c) Stockholm to Jodrell Bank and (d) Jodrell Bank to Stockholm. 15

16 Table 1 gives the delay and standard deviation of inter-packet delay for four BoD routes. End hosts Requested interpacket delay (µs) Mean inter-packet delay (µs) Mode inter-packet delay (µs) EXPReS1 EXPReS Standard deviation of inter-packet delay (µs) EXPReS2 EXPReS1 EXPReS2 jbnexpres2 jbnexpres2 EXPReS2 Table 1. Mean and mode inter-packet delay and standard deviation of inter-packet delay for 1, 2 and 4 Gb/s streams for several BoD routes hour tests UDPmon tests were run between the jbnexpres2 and EXPReS2 hosts for 24 hours in each direction. Data were transmitted at 4Gb/s and statistics reported every 10 seconds. Table 2 summarises the results. Sending host Receiving host EXPReS2 jbnexpres2 jbnexpres2 EXPReS2 Number of packets received Number of packets lost Number of packets reordered Bytes received E E+13 Mean received wire rate (Mbps) Table 2. Results of two 24 hours network tests from Jodrell Bank to Stockholm. 9.5 Discussion (AutoBAHN tests) The expected shape of the plot of inter-packet spacing versus received wire rate has been described elsewhere [12]. Briefly, there is a region of constant received wire rate for a range of inter-packet spacings (a flat horizontal line extending from 0 microseconds inter-packet spacing). At a certain value of inter-packet spacing, which is dependent upon the packet size, the value of received wire rate will begin to decrease. The first, flat region, is explained as the network being saturated with data, and therefore data is transmitted at line rate. In the second region, the received wire rate decreases inversely with increasing inter-packet spacing. This region is encountered when the inter-packet spacing creates a received wire rate lower than the line rate. In the first region, packets are sent from the end host above the line speed, therefore there will be packet loss. With increasing inter-packet spacing, the percentage of packets lost decreases linearly until no packets are lost, which corresponds with the junction in the inter-packet spacing versus received wire rate plot. 16

17 The results from sections 9.1 and 9.2 show that for each of the tested BoD links, multiple packet sizes and inter-packet spacings achieve the maximum bandwidth of 4Gb/s. Three of the tests, (a) London (EXPReS1) and Stockholm (EXPReS2), (b) Stockholm (EXPReS2) and London (EXPReS1), and (c) Stockholm (EXPReS2) and Jodrell Bank (jbnexpres2), show that for the smaller packet sizes, line rate was not achieved (Figure 7, panels (a), (b) and (c)). The corresponding packet loss plots shown in Figure 8 reveal no packet loss for the three smallest packet sizes on plots (b) and (c), and non-uniform losses on all other packet sizes on all three tests. Possible explanations for these results are that the end hosts are not capable of processing packets quickly enough due to hardware or software limitations. In particular, the lack of packet losses in Figure 8, panels (b) and (c) would strongly suggest this. These tests were the first to be conducted, and following these results subsequent tests were all performed with the UDPmon process restricted to run on a single CPU core on both end hosts, which seemed to increase throughput at small packet sizes. Therefore we believe that some of the fluctuations in the plots are due to the UDPmon process switching between CPU cores. Another possibility for the observed fluctuations is that the hardware used in the GÉANT PoPs in London and Stockholm are older than the other hardware. The final three tests all show ideal behaviour for both received wire rate and packet loss (panels (d), (e) and (f) from Figures 7 and 8). The machines involved in these tests are recent hardware and this may well be an important factor, as well as having the UDPmon process restricted to a single CPU core. Section 9.3 reports the results of the jitter tests between end hosts over 4 of the routes tested in sections 9.1 and 9.2. The plots in Figure 9 graphically display the spread of delay between the packets, with tightly bunched, tall groupings representing small jitter values. Table 1 presents the mean, mode and standard deviation of packet jitter times. In all results the mean and mode values follow closely the requested interpacket delay times, which shows that the networks and end hosts have little effect on packet timings. The standard deviation of packet jitter for all four routes are smaller than the requested inter-packet spacing. Packet jitter for the routes between Jodrell Bank and Metsähovi are not presented as there was a network outage preventing those tests to be run. The results of the two 24 hour tests given in section 9.4 shows that the results are remarkably similar to one another. In each test almost 4.8 billion packets were transmitted, resulting in 43 Terabytes of data transfer each way. The number of packets lost from Stockholm to Jodrell Bank was zero, giving an error rate of less than one bit in 3.44 x The reverse route showed a small packet loss of 4.38 x 10-7%. Both routes displayed very similar levels of packet reordering and mean received wire rate. Both BoD services are in the pilot phase of their development, rather than production, and as such there are certain limitations. In Stockholm, a single hardware interface for connecting the EXPReS2 host to either BoD network had to be shared. Both groups would have liked to use the resources for a longer period to perform more detailed testing. Competing for this single, mutually exclusive resource resulted in less network test time available for each BoD network, however following the ethos of equitable sharing of common resources in BoD, this will inevitably be a situation faced in the future. 17

18 9.6 NSI network performance tests Tests using NSI were performed from each of the end-hosts (iperf1.amsterdam1 in Amsterdam, EXPReS2 in Stockholm and NEXPReS in Onsala) to JIVE. For each of the tests, we generated traffic at 4Gb/s using Jumbo frames of (nearly) 9000 bytes. The iperf application was used for these tests, creating well-spaced packets which closely matches the real e-vlbi traffic. Table 3 shows the results of the tests between OSO and JIVE, and Figure 10 the throughput and packet loss as a graph. Sending host Receiving host NEXPReS (OSO) SFXC-e0 (JIVE) Number of packets received Number of packets lost 545 ( %) Number of packets reordered 1 Bytes received E+12 Mean received wire rate (Mbps) Table 3: Results from a 50 minute 4Gb/s test between OSO and JIVE Figure 10: Results from a 50 minute test running on a path created by NSI. Note that in Figure 10, the Y-axis is magnified in order to show the few events of packet loss. The graph shows that there are only 3 seconds that have packet loss, all other seconds are error-free. The three loss events are resp. 167, 101 and 201 lost packets for a total of 545, out of packets, which amounts to only 3 lost packets per million transported packets. The actual throughput is slightly higher than the requested 4Gb/s, most likely due to rounding errors in the iperf application. 18

19 10 Conclusion The tests we have performed show that it is indeed now possible to carry e-vlbi traffic at 4Gb/s over international, inter-domain BoD paths. Paths were configured using both NSI and AutoBAHN and both resulted in near-perfect network performance, certainly sufficient to carry out e-vlbi. Although these BoD systems are currently only testbeds, with limited coverage and limited available bandwidth, we hope that this work will encourage the wider adoption of BoD as an international networking service. 19

20 11 Acknowledgements The authors would like to thank all people involved in the setup of the networks for their work, which has made these tests possible. We realize that setting up these networks involved many more people whose names never even came to our attention. Anthony Ryan and colleagues from NetNorthWest, David Salmon, Dave Tinkler and colleagues from JANET, Babatunde Omogbai, Guy Robertst and colleagues from GÉANT, Fredrik Pettai, Henrik Thostrup Jensen, Frank Blankman and colleagues from NORDUnet, Wouter Huisman, Hans Trompert, John MacAuley and colleagues from SURFnet, Jani Myyry and colleagues from Funet, Börje Josefsson and colleagues from SUNET. 20

21 12 References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] The AutoBAHN website, The AutoBAHN Client Portal, NSI Working group website, NSI Protocol Documentation, IEEE 802.1Q, Kershaw, S., Spencer, R. Hughes-Jones, R., Burgess, P., Casey, S., Rushton, A. and Boven, P. Protocols Performance Report - Final Report on Protocols and Network Infrastructure. EXPReS Project, Dallison, S ; Hughes-Jones, R E ; Pezzi, N ; Li, Y. Bringing High-Performance Networking to HEP Users. Computing in High Energy Physics and Nuclear Physics 2004, Interlaken, Switzerland, 27 Sep - 1 Oct 2004, pp.1304 Boven P. (for the EXPReS team); e-vlbi Networking Tricks, Science and Technology of Long Baseline Real-Time Interferometry (8th e-vlbi WS) Casey S., Hughes-Jones R., Spencer R., Strong M., Burgess P., Szomoru A., Greenwood C. VLBI_UDP: An application for transporting VLBI data using the UDP protocol (2010) Future Generation Computer Systems, 26 (1), pp The UDPmon website, Boven, P., Bloemhof, F., A proof-of-concept system for reserving and provisioning on-demand networking capacity (NEXPReS D6.01), Spencer R.E., Hughes-Jones R., Strong M., Casey S., Rushton A., Burgess P., Kershaw S., Greenwood C. The Role of ESLEA in the development of e-vlbi. (2010) Future Generation Computer Systems, 26 (1), pp

D Demonstration of integrated BoD testing and validation

D Demonstration of integrated BoD testing and validation D6.06 - Demonstration of integrated BoD testing and validation Deliverable Number 06 WP # 6 Date 2012-12-30 Authors Jimmy Cullen, The University of Manchester Fedde Bloemhof, JIVE Paul Boven, JIVE Ralph

More information

PoS(11th EVN Symposium)112

PoS(11th EVN Symposium)112 Arpad Szomoru 1 Joint Institute for VLBI in Europe P.O. Box 2, 7990 AA Dwingeloo, the Netherlands E-mail: szomoru@jive.nl Paul Boven Joint Institute for VLBI in Europe P.O. Box 2, 7990 AA Dwingeloo, the

More information

Commissioning and Using the 4 Gigabit Lightpath from Onsala to Jodrell Bank

Commissioning and Using the 4 Gigabit Lightpath from Onsala to Jodrell Bank Commissioning and Using the 4 Gigabit Lightpath from Onsala to Jodrell Bank 1 DANTE City House,126-13 Hills Road, Cambridge CB2 1PQ, UK And Visiting Fellow, School of Physics and Astronomy, The University

More information

Trans-Atlantic UDP and TCP network tests

Trans-Atlantic UDP and TCP network tests , Paul Burgess, Richard Hughes-Jones, Stephen Kershaw, Ralph Spencer and Matthew Strong The University of Manchester Jodrell Bank Observatory Macclesfield Cheshire SK11 9DL UK E-mail: Anthony.Rushton@postgrad.manchester.ac.uk

More information

SURFnet network developments 10th E-VLBI workshop 15 Nov Wouter Huisman SURFnet

SURFnet network developments 10th E-VLBI workshop 15 Nov Wouter Huisman SURFnet SURFnet network developments 10th E-VLBI workshop 15 Nov 2011 Wouter Huisman SURFnet Engine for Innovation Mission To improve higher education and research by promoting, developing and operating a trusted,

More information

inettest a 10 Gigabit Ethernet Test Unit

inettest a 10 Gigabit Ethernet Test Unit 1 DANTE City House, 126-130 Hills Road, Cambridge CB2 1PQ, UK And Visiting Fellow, School of Physics and Astronomy, The University of Manchester, Oxford Rd, Manchester UK E-mail: Richard.Hughes-Jones@dante.net

More information

Test report of transparent local/remote application programming interface for storage elements

Test report of transparent local/remote application programming interface for storage elements NEXPReS is an Integrated Infrastructure Initiative (I3), funded under the European Community's Seventh Framework Programme (FP7/2007-2013) under grant agreement n RI-261525. Test report of transparent

More information

Lighting the Blue Touchpaper for UK e-science - Closing Conference of ESLEA Project The George Hotel, Edinburgh, UK March, 2007

Lighting the Blue Touchpaper for UK e-science - Closing Conference of ESLEA Project The George Hotel, Edinburgh, UK March, 2007 Working with 1 Gigabit Ethernet 1, The School of Physics and Astronomy, The University of Manchester, Manchester, M13 9PL UK E-mail: R.Hughes-Jones@manchester.ac.uk Stephen Kershaw The School of Physics

More information

evlbi Research in the European VLBI Network Steve Parsley, JIVE

evlbi Research in the European VLBI Network Steve Parsley, JIVE evlbi Research in the European VLBI Network Steve Parsley, JIVE parsley@jive.nl evlbi Research at JIVE and in the EVN is centred on the Proof-of-Concept project set up at the beginning of 2003. The project

More information

PoS(EXPReS09)023. e-evn progress. 3 years of EXPReS. Arpad Szomoru 1

PoS(EXPReS09)023. e-evn progress. 3 years of EXPReS. Arpad Szomoru 1 3 years of EXPReS 1 Joint Institute for VLBI in Europe P.O. Box 2, 7990 AA Dwingeloo, The Netherlands E-mail: szomoru@jive.nl During the past years e-vlbi has undergone a remarkable development. At the

More information

PoS(EXPReS09)036. e-vlbi Networking Tricks. Paul Boven, for the EXPReS team Joint Institute for VLBI in Europe (JIVE)

PoS(EXPReS09)036. e-vlbi Networking Tricks. Paul Boven, for the EXPReS team Joint Institute for VLBI in Europe (JIVE) , for the EXPReS team Joint Institute for VLBI in Europe (JIVE) E-mail: boven@jive.nl Real-time e-vlbi is made possible by having access to long-distance, high-performance networks - straight from the

More information

Lessons from EXPReS and the future of e-vlbi. 2nd Terena E2E Workshop, Amsterdam,

Lessons from EXPReS and the future of e-vlbi. 2nd Terena E2E Workshop, Amsterdam, Lessons from EXPReS and the future of e-vlbi 2nd Terena E2E Workshop, Amsterdam, 2009-12-08 What is JIVE? Operate the EVN correlator and support astronomers doing VLBI. A collaboration of the major radioastronomical

More information

AutoBAHN Provisioning guaranteed capacity circuits across networks

AutoBAHN Provisioning guaranteed capacity circuits across networks AutoBAHN Provisioning guaranteed capacity circuits across networks Afrodite Sevasti, GRNET 1 st End-to-end workshop: Establishing lightpaths 1-2 December 2008, TERENA, Amsterdam AutoBAHN is a research

More information

Deploying Standards-based, Multi-domain, Bandwidth-on-Demand

Deploying Standards-based, Multi-domain, Bandwidth-on-Demand Nordic Infrastructure for Research & Education Deploying Standards-based, Multi-domain, Bandwidth-on-Demand Lars Fischer 28 th NORDUnet Conference Uppsala, 23 September 2014 The State of BoD Hybrid networking

More information

TCPDelay: Constant bit-rate data transfer over TCP

TCPDelay: Constant bit-rate data transfer over TCP The University of Manchester E-mail: Stephen.Kershaw@manchester.ac.uk Richard Hughes-Jones The University of Manchester E-mail: R.Hughes-Jones@manchester.ac.uk Transmission Control Protocol (TCP) is a

More information

Integration of Network Services Interface version 2 with the JUNOS Space SDK

Integration of Network Services Interface version 2 with the JUNOS Space SDK Integration of Network Services Interface version 2 with the JUNOS Space SDK Radosław Krzywania, Michał Balcerkiewicz, Bartosz Belter Poznan Supercomputing and Networking Center, ul. Z. Noskowskiego 12/14,

More information

Connectivity Services, Autobahn and New Services

Connectivity Services, Autobahn and New Services Connectivity Services, Autobahn and New Services Domenico Vicinanza, DANTE EGEE 09, Barcelona, 21 st -25 th September 2009 Agenda Background GÉANT Connectivity services: GÉANT IP GÉANT Plus GÉANT Lambda

More information

Next Generation Networking and The HOPI Testbed

Next Generation Networking and The HOPI Testbed Next Generation Networking and The HOPI Testbed Rick Summerhill Director, Network Research, Architecture, and Technologies, Internet2 CANS 2005 Shenzhen, China 2 November 2005 Agenda Current Abilene Network

More information

ORIENT/ORIENTplus - Connecting Academic Networks in China and Europe. Jennifer(Jie) An CERNET Center/Tsinghua University 14 Feb.

ORIENT/ORIENTplus - Connecting Academic Networks in China and Europe. Jennifer(Jie) An CERNET Center/Tsinghua University 14 Feb. ORIENT/ORIENTplus - Connecting Academic Networks in China and Europe Jennifer(Jie) An CERNET Center/Tsinghua University 14 Feb.2012 Connecting Academic Networks in China and Europe ORIENT project provides

More information

Testbeds as a Service Building Future Networks A view into a new GN3Plus Service. Jerry Sobieski (NORDUnet) GLIF Oct 2013 Singapore

Testbeds as a Service Building Future Networks A view into a new GN3Plus Service. Jerry Sobieski (NORDUnet) GLIF Oct 2013 Singapore Testbeds as a Service Building Future Networks A view into a new GN3Plus Service Jerry Sobieski (NORDUnet) GLIF 2013 4 Oct 2013 Singapore From Innovation to Infrastructure Network Innovation requires testing

More information

Grid Integration of Future Arrays of Broadband Radio Telescopes moving towards e-vlbi

Grid Integration of Future Arrays of Broadband Radio Telescopes moving towards e-vlbi Grid Integration of Future Arrays of Broadband Radio Telescopes moving towards e-vlbi Marcin Okoń Poznań Supercomputing and Networking Center, Supercomputing Department INGRID 2007 S. Margherita Ligure,

More information

National R&E Networks: Engines for innovation in research

National R&E Networks: Engines for innovation in research National R&E Networks: Engines for innovation in research Erik-Jan Bos EGI Technical Forum 2010 Amsterdam, The Netherlands September 15, 2010 Erik-Jan Bos - Chief Technology Officer at Dutch NREN SURFnet

More information

Optical Networking Activities in NetherLight

Optical Networking Activities in NetherLight Optical Networking Activities in NetherLight TERENA Networking Conference 2003 Zagreb, May 19-22, 2003 Erik Radius Manager Network Services, SURFnet Outline NetherLight What is it Why: the rationale From

More information

EVN/JIVE Technical developments. Arpad Szomoru, JIVE

EVN/JIVE Technical developments. Arpad Szomoru, JIVE EVN/JIVE Technical developments Arpad Szomoru, JIVE Correlator upgrades New(ish) data output platform Solaris AMD server with large raid array Fully interchangeable with control computers Re-circulation

More information

GÉANT Support for Research Within and Beyond Europe

GÉANT Support for Research Within and Beyond Europe GÉANT Support for Research Within and Beyond Europe TNC 2003, Zagreb 19-22 May Marian Garcia, Operations Manager DANTE AGENDA GÉANT Global Research Connectivity Europe, North America and other regions

More information

QoS Experience on European Backbone

QoS Experience on European Backbone QoS Experience on European Backbone TNC 2003, Zabgreb (Croatia), 20-05-2003 Nicolas Simar, Network Engineer DANTE Agenda GÉANT Services on GÉANT Premium IP Less than Best Effort Queuing on GÉANT and status

More information

GÉANT Open Service Description. High Performance Interconnectivity to Support Advanced Research

GÉANT Open Service Description. High Performance Interconnectivity to Support Advanced Research GÉANT Open Service Description High Performance Interconnectivity to Support Advanced Research Issue Date: 20 July 2015 GÉANT Open Exchange Overview Facilitating collaboration has always been the cornerstone

More information

Data Gathering in Optical Networks with the TL1 Toolkit

Data Gathering in Optical Networks with the TL1 Toolkit Data Gathering in Optical Networks with the TL1 Toolkit Ronald van der Pol 1 and Andree Toonk 2 1 SARA Computing & Networking Services, Kruislaan 415, 1098 SJ Amsterdam, The Netherlands rvdp@sara.nl http://nrg.sara.nl/

More information

VLBI Participation in Next Generation Network Development

VLBI Participation in Next Generation Network Development VLBI Participation in Next Generation Network Development evlbi Workshop Haystack Observatory 9 th April 2002 Steve Parsley parsley@jive.nl Agenda evlbi Network requirements Existing networks and future

More information

Connecting the e-infrastructure chain

Connecting the e-infrastructure chain Connecting the e-infrastructure chain Internet2 Spring Meeting, Arlington, April 23 rd, 2012 Peter Hinrich & Migiel de Vos Topics - About SURFnet - Motivation: Big data & collaboration - Collaboration

More information

DICE Diagnostic Service

DICE Diagnostic Service DICE Diagnostic Service Joe Metzger metzger@es.net Joint Techs Measurement Working Group January 27 2011 Background Each of the DICE collaborators are defining and delivering services to their users. A

More information

GÉANT3 Services. Ann Harding, SWITCH TNC Connectivity and Monitoring Services by and for NRENs. connect communicate collaborate

GÉANT3 Services. Ann Harding, SWITCH TNC Connectivity and Monitoring Services by and for NRENs. connect communicate collaborate GÉANT3 Services Connectivity and Monitoring Services by and for NRENs Ann Harding, SWITCH TNC 2010 Positioning Services GÉANT benefits multiple disciplines, from Big Science projects such as the Large

More information

Random Neural Networks for the Adaptive Control of Packet Networks

Random Neural Networks for the Adaptive Control of Packet Networks Random Neural Networks for the Adaptive Control of Packet Networks Michael Gellman and Peixiang Liu Dept. of Electrical & Electronic Eng., Imperial College London {m.gellman,p.liu}@imperial.ac.uk Abstract.

More information

Modelling direct application to network bandwidth provisioning for high demanding research applications

Modelling direct application to network bandwidth provisioning for high demanding research applications Modelling direct application to network bandwidth provisioning for high demanding research applications H. Wessing, Y. Yan and M. Berger Research Center COM Technical University of Denmark Bldg. 345V,

More information

CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION

CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION 31 CHAPTER 3 GRID MONITORING AND RESOURCE SELECTION This chapter introduces the Grid monitoring with resource metrics and network metrics. This chapter also discusses various network monitoring tools and

More information

LAN Based Radio Synchronization

LAN Based Radio Synchronization LAN Based Radio Synchronization Introduction The latest generation of Spectralink IP-DECT base stations supports to use Local Area Network (LAN) for synchronization of the DECT radios when running software

More information

Anthony Rushton. University of Manchester

Anthony Rushton. University of Manchester Application of VSI-E Anthony Rushton University of Manchester VSI-E Transfer Protocol (VTP) This work attemps to comply with the draft VSI-E standard Rev. 2.7 VSI-E Tranfer protocol (VTP - version 1.0.P)

More information

Access to the Web. Coverage. Basic Communication Technology. CMPT 165: Review

Access to the Web. Coverage. Basic Communication Technology. CMPT 165: Review Access to the Web CMPT 165: Review Tamara Smyth, tamaras@cs.sfu.ca School of Computing Science, Simon Fraser University December 5, 2011 Access to the Web requires: a computer (of some kind) a connection

More information

GÉANT and other projects Update

GÉANT and other projects Update GÉANT and other projects Update Matthew Scott, General Manager DANTE Internet2 Spring Member Meeting 28 April 2009 Matthew.Scott@dante.net A refresher what is GÉANT? Local campus networks link to national

More information

Deliverable 7.3. Workflow Manager. Poznao Supercomputing and Networking Center

Deliverable 7.3. Workflow Manager. Poznao Supercomputing and Networking Center Deliverable 7.3 Workflow Manager Poznao Supercomputing and Networking Center Table of contents 1 Introduction... 3 2 E-VLBI Correlation Platform... 3 2.1 Liferay introduction... 3 2.2 NEXPReS model...

More information

GÉANT : e-infrastructure connectivity for the data deluge

GÉANT : e-infrastructure connectivity for the data deluge GÉANT : e-infrastructure connectivity for the data deluge Richard Hughes-Jones, Maria Minaricova & Vincenzo Capone DANTE BioMedBridges Workshop, Hinxton, 15-16 May 2014 What is GÉANT? High-bandwidth, high-performance

More information

Southern Polytechnic State University Spring Semester 2009

Southern Polytechnic State University Spring Semester 2009 Southern Polytechnic State University Spring Semester 2009 ECET 4840 Laboratory Exercises 9: Router Queuing Configuration and Testing Objective: Students will investigate the effect of varying link capacity

More information

White paper ETERNUS Extreme Cache Performance and Use

White paper ETERNUS Extreme Cache Performance and Use White paper ETERNUS Extreme Cache Performance and Use The Extreme Cache feature provides the ETERNUS DX500 S3 and DX600 S3 Storage Arrays with an effective flash based performance accelerator for regions

More information

Chin Guok, ESnet. Network Service Interface Signaling and Path Finding

Chin Guok, ESnet. Network Service Interface Signaling and Path Finding GFD-I.234 NSI-WG nsi-wg@ogf.org John MacAuley, ESnet Chin Guok, ESnet Guy Roberts, GEANT August 8, 2017 Network Service Interface Signaling and Path Finding Status of This Document This document provides

More information

WP7. Mark Kettenis April 18, Joint Institute for VLBI in Europe

WP7. Mark Kettenis April 18, Joint Institute for VLBI in Europe WP7 Mark Kettenis kettenis@jive.nl Joint Institute for VLBI in Europe April 18, 2011 Mark Kettenis kettenis@jive.nl (JIVE) NEXPReS WP7 April 18, 2011 1 / 16 Goals Create an automated, distributed correlator

More information

High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK

High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK High bandwidth, Long distance. Where is my throughput? Robin Tasker CCLRC, Daresbury Laboratory, UK [r.tasker@dl.ac.uk] DataTAG is a project sponsored by the European Commission - EU Grant IST-2001-32459

More information

Heterogeneous Interconnection between SDN and Layer2 Networks based on NSI

Heterogeneous Interconnection between SDN and Layer2 Networks based on NSI Heterogeneous Interconnection between SDN and Layer2 Networks based on NSI Ta-Yuan Chou, Wun-Yuan Huang, Hui-Lan Lee, Te-Lung Liu, Joe Mambretti*, Jim Hao Chen*, Fei Yeh* National Center for High-Performance

More information

Cisco XR Series Service Separation Architecture Tests

Cisco XR Series Service Separation Architecture Tests Cisco XR 12000 Series Service Separation Architecture Tests Introduction In April 2005, Cisco Systems released the XR 12000 Series routers bringing the functionality of IOS XR to the edge of next generation

More information

GÉANT Open Service Description. High Performance Interconnectivity to Support Advanced Research

GÉANT Open Service Description. High Performance Interconnectivity to Support Advanced Research GÉANT Open Service Description High Performance Interconnectivity to Support Advanced Research Issue Date: 1 November 2017 GÉANT Open Exchange Overview Facilitating collaboration has always been the cornerstone

More information

Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes

Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes Impact of bandwidth-delay product and non-responsive flows on the performance of queue management schemes Zhili Zhao Dept. of Elec. Engg., 214 Zachry College Station, TX 77843-3128 A. L. Narasimha Reddy

More information

Real-Time E-VLBI with the SFXC Software Correlator. Mark Kettenis Aard Keimpema JIVE

Real-Time E-VLBI with the SFXC Software Correlator. Mark Kettenis Aard Keimpema JIVE Real-Time E-VLBI with the SFXC Software Correlator Mark Kettenis kettenis@jive.nl Aard Keimpema keimpema@jive.nl JIVE SFXC SFXC = S? FX Correlator Where S? used to mean Super Uses MPI to distribute/parallelize

More information

31270 Networking Essentials Focus, Pre-Quiz, and Sample Exam Answers

31270 Networking Essentials Focus, Pre-Quiz, and Sample Exam Answers 31270 Networking Essentials Focus, Pre-Quiz, and Sample Exam Answers CONTENTS Focus Questions... 2 Chapter 1: Explore the Network... 2 Chapter 2: Configure a Network Operating System... 5 Chapter 3: Network

More information

EXPReS SA1: at the half-way mark. Arpad Szomoru JIVE

EXPReS SA1: at the half-way mark. Arpad Szomoru JIVE EXPReS SA1: at the half-way mark Arpad Szomoru JIVE Outline Progress in 2007 Going intercontinental The next 18 months Slide #2 e-vlbi science/test runs 17 e-vlbi science projects accepted since 2006:

More information

Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures

Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures Mauro Campanella - GARR FP7 Future Networks Concertation Meeting Brussels, March 11th, 2008 FEDERICA

More information

UNIT IV -- TRANSPORT LAYER

UNIT IV -- TRANSPORT LAYER UNIT IV -- TRANSPORT LAYER TABLE OF CONTENTS 4.1. Transport layer. 02 4.2. Reliable delivery service. 03 4.3. Congestion control. 05 4.4. Connection establishment.. 07 4.5. Flow control 09 4.6. Transmission

More information

European VLBI Network

European VLBI Network European VLBI Network Cormac Reynolds, JIVE European Radio Interferometry School, Bonn 12 Sept. 2007 EVN Array 15 dissimilar telescopes Observes 3 times a year (approx 60 days per year) Includes some

More information

Multi Domain Service Architecture for Heterogonous Networks A view from GÉANT 3 - SA2: Task 1

Multi Domain Service Architecture for Heterogonous Networks A view from GÉANT 3 - SA2: Task 1 Multi Domain Service Architecture for Heterogonous Networks A view from GÉANT 3 - SA2: Task 1 Brian Bach Mortensen, NORDUnet 2nd TERENA E2E Workshop on Provisioning E2E services & On-demand Infrastructure

More information

4.2 Virtual Circuit and Datagram Networks

4.2 Virtual Circuit and Datagram Networks 4.2 VIRTUAL CIRCUIT AND DATAGRAM NETWORKS 313 Available bit rate (ABR) ATM network service. With the Internet offering socalled best-effort service, ATM s ABR might best be characterized as being a slightly-better-than-best-effort

More information

The Network Infrastructure for Radio Astronomy in Italy

The Network Infrastructure for Radio Astronomy in Italy The Network Infrastructure for Radio Astronomy in Italy 1 INAF Istituto di Radioastronomia Via Gobetti 101, Bologna, Italy E-mail:m.nanni@ira.inaf.it Franco Mantovani INAF Istituto di Radioastronomia Via

More information

PART IV. Internetworking Using TCP/IP

PART IV. Internetworking Using TCP/IP PART IV Internetworking Using TCP/IP Internet architecture, addressing, binding, encapsulation, and protocols in the TCP/IP suite Chapters 20 Internetworking: Concepts, Architecture, and Protocols 21 IP:

More information

GÉANT Plus Service Description. High Performance Cost-effective Connectivity

GÉANT Plus Service Description. High Performance Cost-effective Connectivity GÉANT Plus Service Description High Performance Cost-effective Connectivity Issue Date: 1 November 2017 GÉANT Plus Overview The GÉANT Plus service is offered on the GÉANT network to all GÉANT National

More information

Introduction to Internetworking

Introduction to Internetworking CHAPTER Introduction to Internetworking Introduction This chapter explains basic internetworking concepts. The information presented here helps readers who are new to internetworking comprehend the technical

More information

SMART Questionnaire. Fields marked with * are mandatory. Introduction

SMART Questionnaire. Fields marked with * are mandatory. Introduction SMART Questionnaire Fields marked with are mandatory. Introduction Dear Sir or Madam, We have been tasked by the European Commission's DG CONNECT to gain a detailed understanding of the different measurement

More information

XenaCompact TM Series

XenaCompact TM Series Series Evaluation report Date: 18 January 2010 Version No.: 1.0 All Rights Reserved NORDUnet A/S Contents 1 Introduction...3 2 XENA Compact evaluation...3 2.1 XENA Compact brief description...3 2.2 Initial

More information

10th International Conference on Networks (ICN 2011) Moderator:

10th International Conference on Networks (ICN 2011) Moderator: Panel @ 10th International Conference on Networks (ICN 2011) Moderator: Stein Gjessing, University of Oslo, Norway Panelists: Börje Josefsson, SUNET, Sweden Stein Gjessing, University of Oslo, Norway Andreas

More information

UNIT 2 TRANSPORT LAYER

UNIT 2 TRANSPORT LAYER Network, Transport and Application UNIT 2 TRANSPORT LAYER Structure Page No. 2.0 Introduction 34 2.1 Objective 34 2.2 Addressing 35 2.3 Reliable delivery 35 2.4 Flow control 38 2.5 Connection Management

More information

Integrated Services. Integrated Services. RSVP Resource reservation Protocol. Expedited Forwarding. Assured Forwarding.

Integrated Services. Integrated Services. RSVP Resource reservation Protocol. Expedited Forwarding. Assured Forwarding. Integrated Services An architecture for streaming multimedia Aimed at both unicast and multicast applications An example of unicast: a single user streaming a video clip from a news site An example of

More information

Sections Describing Standard Software Features

Sections Describing Standard Software Features 27 CHAPTER This chapter describes how to configure quality of service (QoS) by using automatic-qos (auto-qos) commands or by using standard QoS commands. With QoS, you can give preferential treatment to

More information

Network Services Interface. OGF NSI standards development progress:

Network Services Interface. OGF NSI standards development progress: Network Services Interface OGF NSI standards development progress: NSI Framework Doc (spr 11) NSI CS v1.0draft (Aug 11) Feed initial implementation NSI CS v1.0 final (Dec 11) NSI CS v2.0draft, NSI Topology

More information

GÉANT Mission and Services

GÉANT Mission and Services GÉANT Mission and Services Vincenzo Capone Senior Technical Business Development Officer CREMLIN WP2 Workshop on Big Data Management 15 February 2017, Moscow GÉANT - Networks Manages research & education

More information

Network Design Considerations for Grid Computing

Network Design Considerations for Grid Computing Network Design Considerations for Grid Computing Engineering Systems How Bandwidth, Latency, and Packet Size Impact Grid Job Performance by Erik Burrows, Engineering Systems Analyst, Principal, Broadcom

More information

Chapter 4. Routers with Tiny Buffers: Experiments. 4.1 Testbed experiments Setup

Chapter 4. Routers with Tiny Buffers: Experiments. 4.1 Testbed experiments Setup Chapter 4 Routers with Tiny Buffers: Experiments This chapter describes two sets of experiments with tiny buffers in networks: one in a testbed and the other in a real network over the Internet2 1 backbone.

More information

Latency on a Switched Ethernet Network

Latency on a Switched Ethernet Network Page 1 of 6 1 Introduction This document serves to explain the sources of latency on a switched Ethernet network and describe how to calculate cumulative latency as well as provide some real world examples.

More information

Cross Layer Protocol Design. Radio Communication III

Cross Layer Protocol Design. Radio Communication III Cross Layer Protocol Design Radio Communication III The layered world of protocols The ISO OSI model OSI model Introduction» The open systems interconnection reference model (OSI model) describes a layered

More information

GÉANT L3VPN Service Description. Multi-point, VPN services for NRENs

GÉANT L3VPN Service Description. Multi-point, VPN services for NRENs GÉANT L3VPN Service Description Multi-point, VPN services for NRENs Issue Date: 1 November 2017 GÉANT L3VPN Overview The GÉANT L3VPN service offers the National Research and Education Networks (NRENs)

More information

GÉANT IP Service Description. High Performance IP Services to Support Advanced Research

GÉANT IP Service Description. High Performance IP Services to Support Advanced Research GÉANT IP Service Description High Performance IP Services to Support Advanced Research Issue Date: 1 November 2017 GÉANT IP Overview The GÉANT IP service provides high-bandwidth, international Internet

More information

II. Principles of Computer Communications Network and Transport Layer

II. Principles of Computer Communications Network and Transport Layer II. Principles of Computer Communications Network and Transport Layer A. Internet Protocol (IP) IPv4 Header An IP datagram consists of a header part and a text part. The header has a 20-byte fixed part

More information

An evaluation tool for Wireless Digital Audio applications

An evaluation tool for Wireless Digital Audio applications An evaluation tool for Wireless Digital Audio applications Nicolas-Alexander Tatlas 1, Andreas Floros 2, and John Mourjopoulos 3 1 Audiogroup, Electrical Engineering and Computer Technology Department,

More information

Zhengyang Liu University of Virginia. Oct 29, 2012

Zhengyang Liu University of Virginia. Oct 29, 2012 SDCI Net: Collaborative Research: An integrated study of datacenter networking and 100 GigE wide-area networking in support of distributed scientific computing Zhengyang Liu University of Virginia Oct

More information

Towards Network Awareness in LHC Computing

Towards Network Awareness in LHC Computing Towards Network Awareness in LHC Computing CMS ALICE CERN Atlas LHCb LHC Run1: Discovery of a New Boson LHC Run2: Beyond the Standard Model Gateway to a New Era Artur Barczyk / Caltech Internet2 Technology

More information

Appropriate Layer-2 Interconnection Between IXPs

Appropriate Layer-2 Interconnection Between IXPs Appropriate Layer-2 Interconnection Between IXPs Keith Mitchell NANOG31, San Francisco 24/25 th May 2004 Layer-2 Interconnect Typical scenarios faced by IXP operators today: ISPs conserving router ports

More information

GRIDS INTRODUCTION TO GRID INFRASTRUCTURES. Fabrizio Gagliardi

GRIDS INTRODUCTION TO GRID INFRASTRUCTURES. Fabrizio Gagliardi GRIDS INTRODUCTION TO GRID INFRASTRUCTURES Fabrizio Gagliardi Dr. Fabrizio Gagliardi is the leader of the EU DataGrid project and designated director of the proposed EGEE (Enabling Grids for E-science

More information

University of Amsterdam

University of Amsterdam The road to optical networking www.science.uva.nl/~delaat www.science.uva.nl/research/air Cees de Laat University of Amsterdam SURFnet λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ λ Know the user #

More information

Deliverable D6.2 VNF/SDN/EPC: integration and system testing

Deliverable D6.2 VNF/SDN/EPC: integration and system testing 5G Communication with a Heterogeneous, Agile Mobile network in the Pyeongchang Winter Olympic CompetitioN Grant agreement n. 723247 Deliverable D6.2 VNF/SDN/EPC: integration and system Date of Delivery:

More information

QoS Configuration. Page 1 of 13

QoS Configuration. Page 1 of 13 QoS Configuration Page 1 of 13 Contents Chapter 1 QoS Configuration...3 1.1 Brief Introduction to QoS...3 1.1.1 Traffic...3 1.1.2 Traffic Classification... 3 1.1.3 Priority...4 1.1.4 Access Control List...

More information

Latency on a Switched Ethernet Network

Latency on a Switched Ethernet Network FAQ 07/2014 Latency on a Switched Ethernet Network RUGGEDCOM Ethernet Switches & Routers http://support.automation.siemens.com/ww/view/en/94772587 This entry is from the Siemens Industry Online Support.

More information

Deliverable D7.4 Field Validation Test Report

Deliverable D7.4 Field Validation Test Report VIVALDI PROJECT DOCUMENT WORK PACKAGE 7 Deliverable D7.4 Field Validation Test Report Revision 1.2 Project full title: Advancing interactive Broadband satellite access by optimal convergence of session

More information

Sections Describing Standard Software Features

Sections Describing Standard Software Features 30 CHAPTER This chapter describes how to configure quality of service (QoS) by using automatic-qos (auto-qos) commands or by using standard QoS commands. With QoS, you can give preferential treatment to

More information

FEDERICA Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures

FEDERICA Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures FEDERICA Federated E-infrastructure Dedicated to European Researchers Innovating in Computing network Architectures Mauro Campanella - GARR Joint Techs Workshop / APAN Honolulu, January 23 2008 FEDERICA

More information

ThousandEyes for. Application Delivery White Paper

ThousandEyes for. Application Delivery White Paper ThousandEyes for Application Delivery White Paper White Paper Summary The rise of mobile applications, the shift from on-premises to Software-as-a-Service (SaaS), and the reliance on third-party services

More information

MPA (Marker PDU Aligned Framing for TCP)

MPA (Marker PDU Aligned Framing for TCP) MPA (Marker PDU Aligned Framing for TCP) draft-culley-iwarp-mpa-01 Paul R. Culley HP 11-18-2002 Marker (Protocol Data Unit) Aligned Framing, or MPA. 1 Motivation for MPA/DDP Enable Direct Data Placement

More information

416 Distributed Systems. Networks review; Day 1 of 2 Jan 5 + 8, 2018

416 Distributed Systems. Networks review; Day 1 of 2 Jan 5 + 8, 2018 416 Distributed Systems Networks review; Day 1 of 2 Jan 5 + 8, 2018 1 Distributed Systems vs. Networks Low level (c/go) Run forever Support others Adversarial environment Distributed & concurrent Resources

More information

1 Energy Efficient Protocols in Self-Aware Networks

1 Energy Efficient Protocols in Self-Aware Networks Energy Efficient Protocols in Self-Aware Networks Toktam Mahmoodi Centre for Telecommunications Research King s College London, London WC2R 2LS, UK Stanford NetSeminar 13 December 2011 1 Energy Efficient

More information

SURFnet6 Integrating the IP and Optical worlds Erik-Jan Bos Director of Network Services SURFnet, The Netherlands TrefPunkt Kiruna, mars 2004

SURFnet6 Integrating the IP and Optical worlds Erik-Jan Bos Director of Network Services SURFnet, The Netherlands TrefPunkt Kiruna, mars 2004 SURFnet6 Integrating the IP and Optical worlds Erik-Jan Bos Director of Network Services SURFnet, The Netherlands TrefPunkt Kiruna, 30-31 mars 2004 Presentation contents SURFnet and GigaPort: The environment

More information

Cisco SAN Analytics and SAN Telemetry Streaming

Cisco SAN Analytics and SAN Telemetry Streaming Cisco SAN Analytics and SAN Telemetry Streaming A deeper look at enterprise storage infrastructure The enterprise storage industry is going through a historic transformation. On one end, deep adoption

More information

The European DataGRID Production Testbed

The European DataGRID Production Testbed The European DataGRID Production Testbed Franck Bonnassieux CNRS/UREC ENS-Lyon France DataGrid Network Work Package Manager Franck.Bonnassieux@ens-lyon.fr Presentation outline General DataGrid project

More information

SaaS Providers. ThousandEyes for. Summary

SaaS Providers. ThousandEyes for. Summary USE CASE ThousandEyes for SaaS Providers Summary With Software-as-a-Service (SaaS) applications rapidly replacing onpremise solutions, the onus of ensuring a great user experience for these applications

More information

MASERGY S MANAGED SD-WAN

MASERGY S MANAGED SD-WAN MASERGY S MANAGED New Performance Options for Hybrid Networks Business Challenges WAN Ecosystem Features and Benefits Use Cases INTRODUCTION Organizations are leveraging technology to transform the way

More information

AT&T Collaborate TM. Network Assessment Tool

AT&T Collaborate TM. Network Assessment Tool AT&T Collaborate TM Network Assessment Tool 2016 AT&T Intellectual Property. All rights reserved. AT&T, Globe logo, Mobilizing Your World and DIRECTV are registered trademarks and service marks of AT&T

More information

GÉANT perspective of Virtual Networks and Implementation

GÉANT perspective of Virtual Networks and Implementation GÉANT perspective of Virtual Networks and Implementation NORDUnet 2002 Conference Copenhagen 15th of April 2002 Agnès Pouélé, Network Engineer DANTE GÉANT perspective of Virtual Networks and implementation

More information