[13:41:59] --- rik wade has joined
[15:08:52] --- mlshore has joined
[15:22:23] --- rik wade has left: Replaced by new connection
[15:22:24] --- rik wade has joined
[15:22:28] --- sureshkr has joined
[15:25:57] --- dlpartain has joined
[15:30:47] --- sureshkr has left: Replaced by new connection
[15:30:47] --- sureshkr has joined
[15:30:47] --- sureshkr has left
[15:31:54] --- Suresh Krishnan has joined
[15:35:39] <Suresh Krishnan> When is the DoS attack analysis presentation starting?
[15:37:23] <rik wade> doesn't seem to be a scribe assigned.
[15:39:36] <dlpartain> too bad...
[15:55:46] --- dlpartain has left
[16:04:12] <Suresh Krishnan> prevent dos attacks
[16:04:18] <Suresh Krishnan> previous approach
[16:04:23] <Suresh Krishnan> use tcp for contrl
[16:04:29] <Suresh Krishnan> udp for data
[16:04:56] <Suresh Krishnan> new approach
[16:05:05] <Suresh Krishnan> use separate tcp channels for data/control
[16:05:44] <Suresh Krishnan> to simulate DoS attack use a packet generator connected to the FE
[16:06:09] <Suresh Krishnan> control - forwarding connection bandwidth is 10Mbps
[16:07:32] <Suresh Krishnan> control data rate drops dramatically at a certain threshold of data rates
[16:09:10] <Suresh Krishnan> the tests were done using two pthreads one for control and one for data
[16:10:34] <Suresh Krishnan> using tcp for both data and control channel creates a better worst case scenario for control messages
[16:11:21] <Suresh Krishnan> instead of zero throughput at peak (100%) data rate the control rate increases to abot 80% of max control rate
[16:12:12] <Suresh Krishnan> alternate approach : use tcp for control and bccp for data?
[16:12:24] <Suresh Krishnan> experiment was not successful
[16:31:56] --- rik wade has left
[16:39:00] --- mlshore has left: Disconnected
[17:17:19] --- Suresh Krishnan has left: Disconnected
[17:18:04] --- Suresh Krishnan has joined
[17:19:10] --- Suresh Krishnan has left