...
Code Block |
---|
486cottrell@pinger:~$sudo nping -p 80 -c 2 -6 --tcp-connect 2001:da8:270:2018:f816:3eff:fef3:bd3 Starting Nping 0.5.51 ( http://nmap.org/nping ) at 2018-01-13 15:21 PST SENT (0.0021s) Starting TCP Handshake > 2001:da8:270:2018:f816:3eff:fef3:bd3:80 RECV (0.1679s) Handshake with 2001:da8:270:2018:f816:3eff:fef3:bd3:80 completed SENT (1.0041s) Starting TCP Handshake > 2001:da8:270:2018:f816:3eff:fef3:bd3:80 RECV (1.1692s) Handshake with 2001:da8:270:2018:f816:3eff:fef3:bd3:80 completed Max rtt: 165.789ms | Min rtt: 165.073ms | Avg rtt: 165.431ms TCP connection attempts: 2 | Successful connections: 2 | Failed: 0 (0.00%) Tx time: 1.00329s | Tx bytes/s: 159.48 | Tx pkts/s: 1.99 Rx time: 1.16836s | Rx bytes/s: 68.47 | Rx pkts/s: 1.71 Nping done: 1 IP address pinged in 1.17 seconds |
Together with this we , we could use the list of PingER perfSONAR hosts that respond to pings. Note that sometimes pings are blocked to a host but TCP port 80 packets work, e.g. adl-a-ext1.aarnet.net.au (202.158.195.68).
Project
We wrote a script ping-vs-tcp.pl. * for analyzing the PingER nodes, and another ping-vs-tcp-ps.pl for analyzing perfSONAR nodes. For more information see here.
Code Block |
---|
Usage: ping-vs-tcp.pl [opts] Opts: -v-help print this USAGE information --Ddebug debug_level (default=0) -p-prot protocol (6 or '') (default '') --aport application port (default = 80) --ccount count of pings or npings to be sent (default = 10) Function: Ping the host provided in %NODE_DETAILS (the PingER database of hosts) For each host it gets the IP address either from NODE_DETAILS (IPv4) or using the dig command (IPv6). It then Pings and npings the host and gathers the min, average, maximum RTTs and losses and reports them to STDOUT., together with a time stamp and host information such as name, IP address, countrycontry, region etc. Externals: Requires nping (requires root/sudo privs), dig Input: It gets information on the PingER hosts from %NODE_DETAILS using: wget to get the required file from http://www-iepm.slac.stanford.edu/pinger/pingerworld/slaconly-nodes.cf. The required file is saved in in /tmp with a unique name /tmp/nodes-34589020.cf (where the number is based on the process number). Examples: ping-vs-tcp.pl ping-vs-tcp.pl --pprot 6 -a-port 22 --ccount 10 | tee pg-v4-nd.txt ping-vs-tcp.pl -v -help Hint: To turn the output into a real csv file do something like: grep warning pg-v4-db.txt > pg-v4-nd.csv Version=0.4, 1/28/2018, Cottrell |
The output is in comma separated value format and is exported to Excel (one file for IPv4, one for IPv6) where it is analyzed to look at the histograms of AverageAvg(ping_-rtt) -Averageand Avg(nping_-rtt) etc. The script was run from a host (pinger.slac.stanford.edu) located st at SLAC in Northern California in the San Francisco Bay Area. It was run multiple times to ensure the behaviour behavior was not dependent on time of day or day of week etc. To ensure the selected target hosts were not unique in some respect we also obtained a list of target hosts from the perfSONAR project , and repeated the analysis for them. Finally, to ensure there was not something unique about making the measurements from SLAC, we repeated the measurements to the same sets of targets from MAs in China, Malaysia, and Thailand. For more information on the
IPv4 measurements
For the IPv4 data measurements using the list of PingER hosts, for the IPv4 hosts the script ran for about 6 hours and 45 minutes. The distribution histograms shown below show that the agreement between ping and nping average RTT is good:
...
- Histogram of difference in Avg(ping-RTT)-Avg(nping-RTT) for IPv6 measured from China
- Histogram of the relative difference between ping and nping RTTs measured from China
- Cumulative Distribution Frequency of the Avg(ping-RTT) and Avg(nping-RTT) with the differences.
The maximum(abs(Difference))=D is 0.031. We use the Kolmogorov Smirnov two sample test. for the number of observations (N), c(alpha) > D/sqrt(2N/N*N), where alpha indicates that the probability of rejecting the null hypothesis (the null hypothesis says there is no relationship between the two measured phenomena) is 85%.
Also see http://www-iepm.slac.stanford.edu/~cottrell/pinger/synack/ping-vs-synack.html and Synack and https://link.springer.com/content/pdf/10.1007%2F978-3-540-25969-5_14.pdf