[liboping.git] / src / mans / oping.pod

=head1 NAME

oping - send ICMP ECHO_REQUEST to network hosts

=head1 SYNOPSIS

B<oping> [B<-4> | B<-6>] [B<-c> I<count>] [B<-i> I<interval>] I<host> [I<host> [I<host> ...]]

B<oping> [B<-4> | B<-6>] [B<-c> I<count>] [B<-i> I<interval>] B<-f> I<filename>

B<noping> [B<-4> | B<-6>] [B<-c> I<count>] [B<-i> I<interval>] I<host> [I<host> [I<host> ...]]

B<noping> [B<-4> | B<-6>] [B<-c> I<count>] [B<-i> I<interval>] B<-f> I<filename>

=head1 DESCRIPTION

B<oping> uses ICMPv4 or ICMPv6 ECHO_REQUEST packets to measure a hosts
reachability and the network latency. In contrast to the original L<ping(8)>
utility B<oping> can send ICMP packets to multiple hosts in parallel and wait
for all ECHO_RESPONSE packets to arrive. In contrast to the B<fping> utility
(URL is listed in L<"SEE ALSO">) B<oping> can use both, IPv4 and IPv6
transparently and side by side.

B<noping> is an ncurses-based front-end to I<liboping> which displays ping
statistics online and highlights aberrant round-trip times if the terminal
supports colors.

=head1 OPTIONS

=over 4

=item B<-4>

Force the use of IPv4. 

=item B<-6>

Force the use of IPv6.

=item B<-c> I<count>

Send (and receive) I<count> ICMP packets, then stop and exit.

=item B<-i> I<interval>

Send one ICMP packet (per host) each I<interval> seconds. This can be a
floating-point number to specify sub-second precision.

=item B<-t> I<ttl>

Set the IP Time to Live to I<ttl>. This must be a number between (and
including) 1E<nbsp>andE<nbsp>255. If omitted, the value B<64> is used.

=item B<-I> I<address>

Set the source address to use. You may either specify an IP number or a
hostname. You B<cannot> pass the interface name, as you can with GNU's
L<ping(8)> - use the B<-D> option for that purpose.

=item B<-D> I<interface name>

Set the outgoing network device to use.

=item B<-f> I<filename>

Instead of specifying hostnames on the command line, read them from
I<filename>. If I<filename> is B<->, read from C<STDIN>.

If I<oping> is installed with the SetUID-bit, it will set the effective UID to
the real UID before opening the file. In the special (but common) case that
I<oping> is owned by the super-user (UIDE<nbsp>0), this means that privileges
are temporarily dropped before opening the file, in order to prevent users from
reading arbitrary files on the system.

If your system doesn't provide I<saved set-user IDs> (this was an optional
feature before POSIXE<nbsp>2001), the behavior is different because it is not
possible to I<temporarily> drop privileges. The alternative behavior is: If the
real user ID (as returned by L<getuid(2)>) and the effective user ID (as
returned by L<geteuid(2)>) differ, the only argument allowed for this option is
"-" (i.e. standard input).

=item B<-Q> I<qos>

Specify the I<Quality of Service> (QoS) for outgoing packets. This is a
somewhat tricky option, since the meaning of the bits in the IPv4 header has
been revised several times.

The currently recommended method is I<Differentiated Services> which is used in
IPv6 headers as well. There are shortcuts for various predefined
I<per-hop behaviors> (PHBs):

=over 4

=item B<be>

Selects the I<Best Effort> behavior. This is the default behavior.

=item B<ef>

Selects the I<Expedited Forwarding> (EF) per-hop behavior, as defined in
I<RFCE<nbsp>3246>. This PHB is characterised by low delay, low loss and low
jitter, i.e. high priority traffic.

=item B<va>

Selects the I<Voice Admitted> (VA) per-hop behavior, as defined in
I<RFCE<nbsp>5865>. This traffic class is meant for I<Voice over IP> (VoIP)
traffic which uses I<Call Admission Control> (CAC) for reserving network
capacity.

=item  B<af>I<c>I<p>

Selects one of 12E<nbsp>differentiated services code points (DSCPs), which are
organized in four I<classes> with three I<priorities> each. Therefore, I<c>
must be a number betweenE<nbsp>1 throughE<nbsp>4 and I<p> must be a number
betweenE<nbsp>1 throughE<nbsp>3, for example "af13", "af22" and "af41". In each
class, the lower priority number takes precedence over the higher priority
number.

=item B<cs>I<n>

Selects one of the eight I<Class Selector> PHBs. I<n> is a number
betweenE<nbsp>0 throughE<nbsp>7. The class selectors have been defined to be
compatible to the I<Precedence> field in the IPv4 header as defined in
I<RFCE<nbsp>791>. Please note that "cs0" is synonymous to "be".

=back

The old definition of the same bits in the IPv4 header was as I<Type of
Service> (ToS) field, specified in I<RFCE<nbsp>1349>. It defined four possible
values which have appropriate aliases. Please note that this use of the bits is
B<deprecated> and the meaning is limited to IPv4!

=over 4

=item B<lowdelay>

Minimize delay

=item B<throughput>

Maximize throughput

=item B<reliability>

Maximize reliability

=item B<mincost>

Minimize monetary cost

=back

Alternatively, you can also specify the byte manually. You can use either a
decimal number (0-255), a hexadecimal number (0x00-0xff) or an octal number
(00-0377) using the usual "0x" and "0" prefixes for hexadecimal and octal
respectively.

The printed lines will contain information about the QoS field of received
packets if either a non-standard QoS setting was used on outgoing packets or if
the QoS byte of incoming packets is not zero. In other words, the QoS
information is omitted if both, the outgoing and the incoming QoS bytes are
zero. The received byte is always interpreted as
I<Differentiated Services Code Point> (DSCP) and
I<Explicit Congestion Notification> (ECN), even if the deprecated
I<Type of Service> (ToS) aliases were used to specify the bits of outgoing
packets.

=item B<-u>|B<-U>

I<noping only> B<-u> forces UTF-8 output, B<-U> disables UTF-8 output. If
neither is given, the codeset is automatically determined from the locale.

=item B<-g> B<none>|B<prettyping>|B<boxplot>|B<histogram>

I<noping only> Selects the graph to display.

=over 4

=item B<none>

Do not show a graph.

=item B<prettyping>

Show a graph with time on the x-axis in a round-robin fashion, i.e. continue on
the left if the right edge is reached. The y-axis shows the round-trip time.
This is the default.

=item B<boxplot>

Show a I<box plot> where the x-axis, i.e. the width of the window, is the
round-trip time. The entire width of the window it the ping interval, set with
the B<-i> option.

The box is sized so it contains 50% of the replies. The vertical line shows the
median. The whiskers are sized to contain 95% of the replies -- 2.5% below the
whiskers and 2.5% above.

  |----------[#####|##########]--------------------------------------------|
  ^          ^     ^          ^                                            ^
 2.75%      25%   50%        75%                                         97.5%

=item B<histogram>

Show a I<histrogram> of the round-trip times. The width of the window is taken
as round-trip time from 0ms on the left to the I<interval> (the B<-i> option,
default 1000ms) on the right.

The height of the graph is scaled so that the most-used buckets vertically fills
the line. The buckets are colored green up to and including the median, yellow
up to and including the 95th percentile and red for the remainder.

In addition, the histogram is colored, if supported by the terminal. The green
bars include the median (50% of responses), yellow bars include the 95th
percentile and outliers are colored red.

=back

=item B<-P> I<percent>

Configures the latency percentile to report. I<percent> must be a number
between zero and 100, exclusively in both cases. In general, defaults to B<95>.
If B<-c> is given and a number less than 20, this would be the same as the
maximum. In this case the default is chosen so that it excludes the maximum,
e.g. if B<-cE<nbsp>5> is given, the default is I<80>.

The calculated percentile has roughly millisecond precision. If precision is of
importance, read on for a more detailed explanation. In order to calculate the
percentile without keeping all replies in memory, I<oping> divides the
I<interval> (the B<-i> option) in 1000 "buckets". Each bucket counts the number
of packets received in the represented time. That means that the precision
decreases if the interval is increased, because each bucket represents a larger
(fraction of the) response time. The code looks for the first bucket
representing at least I<percent> responses and returns the upper-bound latency
represented by that bucket. Since the represented percentage may be larger than
the configured percentile, this algorithm I<overestimes> the actual percentile
by at most 1000th of I<interval>.

=item B<-Z> I<percent>

If any hosts have a drop rate higher than I<percent>, where I<percent> is a
number between zero and 100 inclusively, exit with a non-zero exit status.
Since it is not possible to have a higher drop rate than 100%, passing this
limit will effectively disable the feature (the default). Setting the option to
zero means that the exit status will only be zero if I<all> replies for I<all>
hosts have been received.

The exit status will indicate the number of hosts with more than I<percent>
packets lost, up to a number of 255 failing hosts.

=back

=head1 COLORS

If supported by the terminal, I<noping> will highlight the round-trip times
(RTT) using the colors green, yellow and red. Green signals RTTs that are in
the "expected" range, yellow marks moderately unusual times and times that
differ a lot from the expected value are printed in red.

The information used to categorize round-trip times is the I<percentile>. RTTs
in the 80th percentile are considered to be "normal" and are printed in green.
RTTs within the 95th percentile are considered "moderately unusual" and are
printed in yellow. RTTs above that are considered to be "unusual" and are
printed in red.

=head1 SEE ALSO

L<ping(8)>, L<http://www.fping.com/>, L<liboping(3)>

=head1 AUTHOR

liboping is written by Florian "octo" Forster E<lt>ff at octo.itE<gt>.
Its homepage can be found at L<http://verplant.org/liboping/>.

Copyright (c) 2005-2011 by Florian "octo" Forster.