<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html><head>
<meta http-equiv="Content-Type"
content="text/html;charset=iso-8859-1"><title>GPSD-NG client driver</title>
<link href="scripts/style.css" type="text/css" rel="stylesheet">
<style type="text/css">
table.dlstable { font-size:85%; }
td.ttf{ font-family:Courier; font-weight:bold; }
</style></head>
<body>
<h3>GPSD NG client driver</h3>
<p>Last update:
<!-- #BeginDate format:En2m -->30-Apr-2015 05:53<!-- #EndDate -->
UTC</p>
<hr>
<h4>Synopsis</h4>
<p>
Address: 127.127.46.<i>u</i><br>
Reference ID: <tt>GPSD</tt><br>
Driver ID: <tt>GPSD_JSON</tt><br>
Serial Port: <tt>/dev/gps<i>u</i></tt> as symlink to the true
device (not used directly; see below)<br>
Features: <tt></tt>
</p>
<!-- --------------------------------------------------------- -->
<br><h4>Description</h4>
<p>
This driver is a client driver to the <i>GPSD</i> daemon, which
over the time became increasingly popular for UN*Xish
platforms. <i>GPSD</i> can manage several devices in parallel,
aggregate information, and acts as a data hub for client
applications. <i>GPSD</i> can also auto-detect and handle PPS
hardware signals on serial ports. Have a look
at <a href="http://www.catb.org/gpsd/">the
<i>GPSD</i> project page</a>.
</p>
<p>
<b>It is important to understand that this driver works best
using a GPS device with PPS support.</b>
</p>
<p>
The GPSD-NG protocol is text based, using JSON notation to
transfer records in form of JSON objects. The driver uses a
TCP/IP connection to <tt>localhost:gpsd</tt> to connect to the
daemon and then requests the GPS
device <tt>/dev/gps<i>u</i></tt> to be watched. (Different clock
units use different devices, and
<i>GPSD</i> is able to give only the relevant information to a clock
instance.)
</p>
<p>
This driver does not expect <i>GPSD</i> to be running or the
clock device to be present <i>a priori</i>; it will try to
re-establish a lost or hitherto unsuccessful connection and will
wait for device to come up in <i>GPSD.</i> There is an initial
10 seconds delay between a connection loss or failed attempt and
the next reconnect attempt; this makes sure that there is no
thrashing on the network layer. If the connection fails again,
an exponential back off is used with an upper limit of
approximately 10 minutes.
</p>
<p>
The overall accuracy depends on the receiver used. The driver
uses the error estimations (95% probability limits) provided by
<i>GPSD</i> to set the clock precision dynamically according to
these readings.
</p>
<p>
The driver needs the VERSION, TPV, PPS, WATCH and TOFF objects
of the <i>GPSD</i> protocol. (Others are quietly ignored.) The
driver can operate without the TOFF objects, which are available
with the <i>protocol</i> version 3.10 and above. (Not to be
confused with the <i>release</i> version of <i>GPSD</i>!)
Running without TOFF objects has a negative impact on the jitter
and offset of the serial timing information; if possible, a
version of <i>GPSD</i> with support for TOFF objects should be
used.
</p>
<p>The acronym <u>STI</u> is used here as a synonym for <i>serial
time information</i> from the data channel of the receiver, no
matter what objects were used to obtain it.
</p>
<!-- --------------------------------------------------------- -->
<br><h4>Naming a Device</h4>
<p>
The <i>GPSD</i> driver uses the same device name as the NMEA
driver, namely <tt>/dev/gps<i>u</i></tt>. There is a simple
reason for that: While the NMEA driver and the <i>GPSD</i>
driver can be active at the same time <b>for different
devices</b>, they cannot access the same device at a
time. Having the same name helps on that. It also eases
migration from using NMEA directly to using <i>GPSD</i>, as no
new links etc need to be created.
</p>
<p>
<i>GPSD</i> is normally started with the device name to access;
it can also be instructed by hot-plug scripts to add or remove
devices from its device pool. Luckily, the symlinks used by the
NMEA driver are happily accepted and used by <i>GPSD</i>; this
makes it possible to use the symlink names as device
identification. This makes the migration from the built-in NMEA
driver a bit easier.
</p>
<p><b>Note:</b> <i>GPSD</i> (as of version 3.10) cannot use kernel
mode PPS on devices that are hot-plugged. This would require to
attach the PPS line discipline to the character special file,
which is not possible when running with root privileges already
dropped. This is not likely to change in the future.
</p>
<!-- --------------------------------------------------------- -->
<br><h4>The 'mode' word</h4>
<p>
A few operation modes can be selected with the mode word.
</p>
<p>
<table border="1" frame="box" rules="all">
<th colspan="3">The Mode Word</th>
<tr> <td>Bits</td><td>Value</td><td>Description</td>
</tr>
<tr> <td rowspan="4"align="center">0..1</td>
<td align="center">0</td>
<td>STI only operation. This mode is affected by the timing
stability of whatever protocol is used between the GPS
device and GPSD.
<br>
Running on STI only is not recommended in general. Possible
use cases include:
<ul>
<li>The receiver does not provide a PPS signal.
<li>The receiver <i>does</i> provide a PPS signal and
the secondary PPS unit is used.
<li>The receiver has a stable serial timing and a proper
fudge can be established.
<li>You have other time sources available and want to
establish a useful fudge value for <tt>time2</tt>.
</ul>
</td>
</tr>
<tr>
<td align="center">1</td>
<td>Strict operation. This mode needs a valid PPS and a
valid STI to combine the absolute time from the STI with
the time stamp from the PPS record. Does not feed clock
samples if no valid PPS+STI pair is available.
<br><br>
This type of operation results in an ordinary clock with a
very low jitter as long as the PPS data is available, but
the clock fails once PPS drops out. This mode is a
possible choice for receivers that provide a PPS signal
most of the time but have an unstable serial timing that
cannot be fudge-compensated.
</td>
</tr>
<tr><td align="center">2</td>
<td>Automatic mode. Tries to operate in strict mode unless
it fails to process valid samples for some time, currently
120s. Then it reverts to STI-only operation until the PPS
is stable again for 40s, when strict mode is engaged
again.
<br><br><b>Important Notice: This is an expiremental
feature!</b><br> Switching between strict and STI-only
mode will cause changes in offset and jitter. Use this
mode only if STI-only works fairly well with your setup,
or if you expect longer dropouts of the PPS signal and
prefer to use STI alone over not getting synchronised at
all.</td>
</tr>
<tr>
<td align="center">3</td>
<td><i>(reserved for future extension, do not use)</i></td>
</tr>
<tr>
<td align="center">2..31</td>
<td colspan="2"><i>(reserved for future extension, do not
use)</i></td>
</tr>
</table>
</p>
<!-- --------------------------------------------------------- -->
<br><h4>Syslog flood throttle</h4>
<p>This driver can create a lot of syslog messages when things go
wrong, and cluttering the log files is frowned upon. So we
attempt to log persistent or recurring errors only once per
hour. On the other hand, when tracking a problem the syslog
flood throttle can get into the way.</p>
<p>Therefore, fudge <i>flag3</i> can be used to <i>disable</i> the
flood throttle at any time; the throttle is engaged by
default. Running with the syslog flood throttle disabled for
lengthy time is not recommended unless the log files are closely
monitored.</p>
<!-- --------------------------------------------------------- -->
<br><h4>PPS secondary clock unit</h4>
<p>Units with numbers ≥128 act as secondary clock unit for the
primary clock unit (u mod 128). A secondary unit processes only
the PPS data from <i>GPSD</i> and needs the corresponding master
unit to work<a href="#fn1" name="fn1bl"><sup>1</sup></a>. Use
the 'noselect' keyword on the primary unit if you are not
interested in its data.
</p><p>The secondary unit employs the usual precautions before
feeding clock samples:</p>
<ul>
<li>The system must be already in a synchronised state.
<li>The system offset must be less than 400ms absolute.
<li>The phase adjustment from the PPS signal must also be less
than 400ms absolute.
</ul>
<p>If fudge flag <tt>flag1</tt> is set for the secondary unit, the
unit asserts the PPS flag on the clock as long as PPS data is
available. This makes the unit eligible as PPS peer and should
only be used if the GPS receiver can be trusted for the quality
of its PPS signal<a href="fn2"
name="fn2bl"><sup>2</sup></a>. The PPS flag gets cleared if no
PPS records can be aquired for some time. The unit also flushes
the sample buffer at this point to avoid the use of stale PPS
data.</p>
<p><b>Attention:</b> This unit uses its own PPS fudge value
which must be set as fudge <tt>time1</tt>. Only the fudge
values <tt>time1</tt> and <tt>flag1</tt> have an impact on secondary
units.</p>
<!-- --------------------------------------------------------- -->
<br><h4>Clockstats</h4>
<p>If flag4 is set when the driver is polled, a clockstats record
is written for the primary clock unit. (The secondary PPS unit
does not provide clock stats on its own.) The first 3 fields are
the normal date, time, and IP address common to all clockstats
records.
</p><p>
<table border="1" frame="box" rules="all">
<th colspan="2">The Clockstats Line</th>
<tr> <td>field</td><td>Description</td> </tr>
<tr>
<td align="center">1</td>
<td>Date as day number since NTP epoch.</td>
</tr><tr>
<td align="center">2</td>
<td>Time as seconds since midnight.</td>
</tr><tr>
<td align="center">3</td>
<td>(Pseudo-) IP address of clock unit.</td>
</tr><tr>
<td align="center">4</td>
<td>Number of received known JSON records since last
poll. The driver knows about TPV, PPS, TOFF, VERSION and
WATCH records; others are silently ignored.
</td>
</tr><tr>
<td align="center">5</td>
<td>Bad replies since last poll. A record is considered
malformed or a bad reply when it is missing vital fields
or the fields contain malformed data that cannot be
parsed.
</td>
</tr><tr>
<td align="center">6</td>
<td>Number of sample cycles since last poll that were
discarded because there was no GPS fix. This is
effectively the number of TPV records with a fix value
< 2 or without a time stamp.
</td>
</tr><tr>
<td align="center">7</td>
<td>Number of serial time information records (TPV or TOFF,
depending on the GPSD version) received since last poll.
</td>
</tr><tr>
<td align="center">8</td>
<td>Number of serial time information records used for
clock samples since the last poll.
</td>
</tr><tr>
<td align="center">9</td>
<td>Number of PPS records received since the last poll.</td>
</tr><tr>
<td align="center">10</td>
<td>Number of PPS records used for clock samples on the
secondary channel since the last poll.
</td>
</tr>
</table>
</p>
<!-- --------------------------------------------------------- -->
<br><h4>Fudge Factors</h4>
<dl>
<dt><tt>time1 <i>time</i></tt></dt>
<dd>Specifies the PPS time offset calibration factor, in seconds
and fraction, with default 0.0.</dd>
<dt><a name="fudgetime2"><tt>time2 <i>time</i></tt></a></dt>
<dd><em>[Primary Unit]</em> Specifies the TPV/TIME time offset
calibration factor, in seconds and fraction, with default
0.0.</dd>
<dt><tt>stratum <i>number</i></tt></dt>
<dd>Specifies the driver stratum, in decimal from 0 to 15, with
default 0.</dd>
<dt><tt>refid <i>string</i></tt></dt>
<dd>Specifies the driver reference identifier, an ASCII string
from one to four characters, with default <tt>GPSD</tt>.</dd>
<dt><tt>flag1 0 | 1</tt></dt><dd><em>[<b>Secondary</b>
Unit]</em> When set, flags the secondary clock unit as a
potential PPS peer as long as good PPS data is available.
</dd>
<dt><tt>flag2 0 | 1</tt></dt>
<dd><em>[Primary Unit]</em> When set, <u>disables</u> the
processing of incoming PPS records. Intended as an aide to
test the effects of a PPS dropout when using automatic mode
(mode 2).
</dd>
<dt><tt>flag3 0 | 1</tt></dt><dd><em>[Primary Unit]</em>
If set, <u>disables</u> the log throttle. Useful when tracking
problems in the interaction between <i>GPSD</i> and <i>NTPD</i>,
since now all error events are logged. Persistent/recurrent
errors can easily fill up the log, so this should only be
enabled during bug hunts.</dd>
<dt><tt>flag4 0 | 1</tt></dt><dd><em>[Primary Unit]</em>
If set, write a clock stats line on every poll cycle.
</dd>
</dl>
<!-- -- footnotes -------------------------------------------- -->
<hr>
<p><a name="fn1" href="#fn1bl"><sup>1</sup>) </a>Data transmission
an decoding is done only once by the primary unit. The decoded
data is then processed independently in both clock units. This
avoids double transmission over two sockets and decoding the
same data twice, but the primary unit is always needed as a
downside of this approach.
</p>
<p><a name="fn2" href="#fn2bl"><sup>2</sup>) </a>The clock driver
suppresses the processing PPS records when the TPV/TIME data
indicates the receiver has no fix. It can also deal with
situations where the PPS signal is not delivered
to <i>GPSD</i>. But once it is available, it is also processed
and used to create samples. If a receiver cannot be trusted for
the precision of its PPS signal, it should not be used to create
a possible PPS peer: These get extra clout and can effectively
become the sole source of input for the control loop. You do not
want to use sloppy data for that.
<hr>
<p>Additional Information</p>
<p><a href="../refclock.html">Reference Clock Drivers</a></p>
<hr>
<script type="text/javascript" language="javascript" src="scripts/footer.txt"></script>
</body></html>