This API provides an R interface to RSSAC002 metrics colllected by the DNS root servers.
This project is dependent on the following R modules to be installed and in the path.
rssac002.R expects that data from
RSSAC002-DATA is
available in ./RSSAC002-data. From ./ type git clone https://github.com/rssac-caucus/RSSAC002-data to retrieve this data,
and git pull https://github.com/rssac-caucus/RSSAC002-data to update it.
The main file containing the following functions written in R.
metricsByDate(path, letters, startDate, endDate, metrics) Return a vector of values ordered by date.
path is the relative or absolute location of the rssac002 data files as downloaded by getYaml.sh.
letters can be a single letter(a), a range(a-c), a list(a,b,g), or a combination of all 3. letters is case inspecific. If letters is not a single letter an aggregate will be returned.
startDate and endDate take the form YYYY-MM-DD, endDate will not be included Example: 2016-02-26
metrics is a vector of requested metrics as strings arranged in a hierarchy to identify a single vector of values. The length of metrics can be either 2 or 3. The first and second entry of metrics are required. The third entry of metrics is optional dependent upon the metric being queried. Examples: c("traffic-sizes", "udp-request-sizes", "16-31"), c("unique-sources", "num-sources-ipv6"), c("rcode-volume", "10")
If metricsByDate(path, letters, startDate, endDate, c('traffic-sizes', $SIZE-TYPE)) Where $SIZE-TYPE == 'udp-request-sizes' || 'udp-response-sizes' || 'tcp-request-sizes' || 'tcp-response-sizes'. metricsByDate() will return a 2 dimensional list of all sizes of $SIZE-TYPE by date. NOTE: In this special-case instantiation metricsByDate() will return a list and not a vector.
maxN(ll, N) Return N most maximum values from a list preserving names and order.
ll is a list of numeric values
N is number of elements to return
perc(ll, r=2) Compute percentage of sum(ll) for each value in a list preserving names and order.
ll is a list of numeric values
r is passed to round() as digits=r
A directory full of examples of increasing complexity.
Q: What does this mean?
Warning in loop_apply(n, do.ply) : Removed 2 rows containing missing values (stat_smooth).
A: We store all values as doubles with the exception of zero. Zero is stored as an integer. What this error likely means is that we interpreted a value in a YAML file that was larger than could be stored in a double. Often this happens in times of DDOS when the unique-sources metric grows exceptionally large. It's an outlyer and unfortunately the R language doesn't have an easy way to define a new type that can hold these large values. My advice is simply to ignore them.