[13:30:19] standby joins the room [14:20:01] standby leaves the room [16:11:38] fenghongyan2009 joins the room [16:11:54] fenghongyan2009 leaves the room [16:45:41] standby joins the room [16:45:59] standby leaves the room [19:56:32] Robert Raszuk joins the room [19:59:41] maciekwojciechowski joins the room [20:01:18] arifumi joins the room [20:03:26] Mat Ford joins the room [20:04:11] vijay.gurbani joins the room [20:04:22] gigix73 joins the room [20:04:43] I am doubling as the notetaker, so others in the room please help if I can't update the slide numbers ... [20:04:57] acassen joins the room [20:04:59] Mythbustering draft discussion right now [20:05:50] slide title P2P Traffic Optimization [20:06:32] shamus joins the room [20:06:55] ndelregno joins the room [20:08:17] Talking about different approaches - shaping, chaching, and peer selection. [20:08:35] Next slide "Peer selection in p2p apps" [20:10:01] Discussed how peer selection is done in p2p applications - agnostic to network topology and depends on tit-for-tat in BitTorrent, for instance. RTT, jitter estimation are other techniques [20:10:54] p2p localization wants to do better than average. Research papers show it is effective, at least in field trials. Don't know how it will be in the wild [20:11:08] Next slide with the LHC cartoon. [20:11:09] sandoche2k joins the room [20:11:47] ivica.rimac joins the room [20:11:48] next slide: title draft-*mythbustering [20:12:21] ivica.rimac leaves the room [20:12:28] This draft takes a look at various issues discussed in many fora. [20:13:41] Next slide, same title as before [20:14:48] ivica.rimac joins the room [20:15:07] This draft needs to be moved in a community effort from this group. Please provide input. [20:15:28] next slide: myths (to date) [20:15:53] nest slide: reduce cross-domain traffic [20:16:44] standby joins the room [20:17:30] next slide: Increased app. performance. [20:17:34] standby is now known as remote [20:18:24] Increased application performance: 4 references show that this is true. 2 references show some disadvantages. [20:20:45] Time running out -- author wants to get a sense from the room on if this work is worth pursuing. [20:21:19] momose joins the room [20:21:47] Stas Khirman says yes on the mic. [20:21:50] donley.chris joins the room [20:21:52] Richard Yang says yes [20:21:57] remote is now known as standby [20:23:31] nkawa20060905 joins the room [20:23:33] Next presentation by Henning. [20:24:04] next slide: overview [20:24:07] standby leaves the room [20:24:09] momose leaves the room [20:24:38] standby joins the room [20:25:09] next slide: attacker motivation [20:25:53] Went through attacker motivation: could be political, financial, or for fun. [20:27:39] next slide: attacker resources. [20:29:36] next slide: attack timing [20:31:31] next slide: p2p for RT. [20:33:03] slide: file sharing vs. real-time [20:35:29] slide: admission control [20:36:29] slide: positioning in overlay [20:36:52] slide: id of malicious pers [20:37:21] slide: RT services are diff. [20:38:17] slide: Attack [20:39:26] slide: Summary [20:43:25] vijay.gurbani leaves the room [20:45:13] donley.chris leaves the room [20:45:43] vijay.gurbani joins the room [20:46:08] Stevens LeBlond presenting now. [20:46:19] slide: two open questions about BT locality [20:46:39] BT = BitTorrent [20:47:18] slide: Methodology [20:49:03] sandoche2k leaves the room [20:50:16] HASIB3CD1A020 joins the room [20:50:16] slide 6 [20:50:17] ruriham joins the room [20:50:18] slide 7 [20:50:19] maciekwojciechowski leaves the room [20:50:34] sandoche2k joins the room [20:51:42] slide 9 [20:52:04] salfare@gmail.com joins the room [20:52:42] Glenn joins the room [20:53:13] slide 10 [20:53:23] salfare@gmail.com leaves the room [20:54:43] slide 12 [20:57:49] slide 14 [20:58:18] alst joins the room [20:59:48] cabo joins the room [21:00:25] Zoran's talk starts now. [21:02:08] slide 2 [21:02:29] slide 3 [21:02:57] standby leaves the room [21:02:59] alst leaves the room: Disconnected [21:03:55] slide 5 [21:05:03] slide 7 [21:10:33] acassen leaves the room [21:12:19] sftcd joins the room [21:12:27] Stas Khirman: Public Tracker Scraping [21:13:36] slide: Tracker stats [21:14:34] salfare@gmail.com joins the room [21:14:54] Scraped three trackers - TPB, Demonoid, and Mininova. TPB has 1.7M torrents, Demonoid has 11K and Mininovas has 4K. [21:16:10] slide: CDF of peer dist. [21:17:29] HASIB3CD1A020 leaves the room [21:17:54] salfare@gmail.com leaves the room [21:18:24] slide: first 1k torrents [21:18:59] slide: normalized density [21:20:23] slide: Geo distribution [21:21:57] slide: AS dist. [21:23:00] slide: AS dist. [21:24:48] Henning: Can the authors of the three works here get together and agree on the definition of a CDF so we can normalize the data. [21:26:53] Richard: All studies so far focused on ASes. Would be good to get data per ISP, for instance. [21:28:16] Richard Yang, Efficient p2p design using in-network data lockers [21:29:29] slide: motivation [21:30:29] We need to focus on how to reduce the traffic on the last mile. For downlink we do not want to reduce the traffic. We are only interested in the uplink, which is usually the bottleneck. [21:33:29] slide: basic idea [21:35:17] slide: basic usage scenario 1 [21:36:16] slide: scenario 2 [21:36:53] slide: technical issues [21:38:02] slide: LAP [21:39:04] slide: lap data request i'face [21:39:52] slide: data locker resource model [21:41:10] slide: efficient locker data storage [21:41:43] slide: integration w/ alto, p4p [21:43:10] slide: evaluation: BT [21:45:31] Q/A session now. [21:47:00] Question: LAP will be implemented at the client level? Richard: Yes. Q: Did you consider solutions that would be transparent to the clients? Richard: No, do you have a solution? Reinaldo: Why would a client upload to a data locker? Richard: Avoid access congestion on last mile. [21:49:32] Christophe Diot, Nano data centeres [21:49:57] Don't do p2p on pcs, but on devices managed by someone like the ISP [21:50:12] Christian Scholz joins the room [21:52:40] slide 4 [21:55:07] slide 5 [21:55:53] slide 6 [22:00:40] slide 8 [22:00:44] slide 10 [22:04:38] gigix73 leaves the room [22:06:41] cabo leaves the room [22:06:46] meeting over. [22:07:01] vijay.gurbani leaves the room [22:07:01] Mat Ford leaves the room [22:07:21] ndelregno leaves the room [22:07:26] shamus leaves the room [22:08:02] yangyguo joins the room [22:08:37] yangyguo leaves the room [22:08:49] sandoche2k leaves the room [22:08:59] Robert Raszuk leaves the room [22:10:40] ivica.rimac leaves the room [22:14:40] sftcd leaves the room [22:17:37] arifumi leaves the room [22:17:46] ruriham leaves the room [22:19:15] Glenn leaves the room [22:19:34] cabo joins the room [22:22:12] donley.chris joins the room [22:22:18] donley.chris leaves the room [22:27:35] Mat Ford joins the room [22:27:42] Christian Scholz leaves the room [22:30:58] nkawa20060905 leaves the room [22:31:59] Mat Ford leaves the room