Experiment 2: DNS lookup

Captured DNS traffic with Wireshark using filter udp.port==53 then exported in PDML format producing a file: dns-capture-20100915a.pdml

Using Emacs filtered out all lines except those containing dns.time fields

data2<-read.table("dns-capture-20100915a-a.txt", header=FALSE)
summary(data2)       V1
   Min.   :0.000710
   1st Qu.:0.000896
   Median :0.001066
   Mean   :0.023868
   3rd Qu.:0.003329
   Max.   :0.389880
summary of DNS lookup data

Transcript

So, if we think about our second experiment that we described in an earlier module, in this case, we're looking at the time to do a domain name system lookup.  We captured that traffic, and we used the tool Wireshark, and we set the filter to be UDP port equals 53, which is the part number for the DNS traffic. And then, we exported the file in the so-called PDML format, and we produced the file in this case "dns-capture-20100915a.pdml" and using emacs, we removed all the lines except those containing the dns.time field, and now I'll put it into a text file.  And so here we can see that we're going to simply read this table in from our text file, it has no header, since we have in this dataframe we can simply say "summary" of the dataframe and violà, we see the minimum value, the maximum value, the first quartile, third quartile, the median, and the mean values. And now, of course, we can run the function foo, which we showed in the earlier module on that dataframe for a particular length, and we can compute all of the other sorts of statistics we would like to be able to compute. Thus making it possible to run one run after another after another run - to and look at the data for different days, different times of the day, etc.  and get our statistical values from that.