Reference http://en.wikipedia.org/wiki/Network_Time_Protocol, http://support.ntp.org/bin/view/Support/NTPRelatedDefinitions and http://tools.ietf.org/html/rfc5905 for more informations.
/diamond/doc/process:
indexdaemon process
Synchronizing a client to a network server consists of several packet exchanges where each exchange is a pair of request and reply. When sending out a request, the client stores its own time (originate timestamp) into the packet being sent. When a server receives such a packet, it will in turn store its own time (receive timestamp) into the packet, and the packet will be returned after putting a transmit timestamp into the packet. When receiving the reply, the receiver will once more log its own receipt time to estimate the travelling time of the packet. The travelling time (delay) is estimated to be half of "the total delay minus remote processing time", assuming symmetrical delays. See http://www.ntp.org/ntpfaq/NTP-s-algo.htm#Q-ALGO-BASIC-SYNC for more informations.
When repeatedly reading the time, the difference may vary almost randomly. The difference of these differences (second derivation) is called jitter. In probability, it is called white phase noise. See http://www.ntp.org/ntpfaq/NTP-s-sw-clocks-quality.htm for more informations.
How often index queries remote time servers in seconds, value always is power of 2. index will adjust this value itself, value will range from minpoll (default 6. 2^6 = 64 seconds) to maxpoll (default 10, 2^10 = 1024 seconds).
Short polling intervals update the parameters frequently and are sensitive to jitter and random errors, also create more traffic in network. Long intervals may require larger corrections with significant errors between the updates. However there seems to be an optimum between those two. For common operating system clocks this value happens to be close to the default maximum polling time, 1024 seconds. See http://www.ntp.org/ntpfaq/NTP-s-algo.htm#Q-ALGO-POLL-BEST for more information.
The value displayed in column reach is octal, and it represents the reachability register. One digit in the range of 0 to 7 represents three bits. The initial value of that register is 0, and after every poll that register is shifted left by one position. If the corresponding time source sent a valid response, the rightmost bit is set.
During a normal startup the registers values are these: 0, 1, 3, 7, 17, 37, 77, 177, 377. If reach is another number, such as 257, in binary is 10101111, saying that two valid responses were not received during the last eight polls. However, the last four polls worked fine.
index uses a hierarchical, semi-layered system of time sources. Each level of this hierarchy is termed a stratum and is assigned a number starting with zero at the top. The number represents the distance from the reference clock and is used to prevent cyclical dependencies in the hierarchy. Stratum is not always an indication of quality or reliability; it is common to find stratum 3 time sources that are higher quality than other stratum 2 time sources. Telecommunication systems use a different definition for clock strata. Only strata 0 to 15 are valid; stratum 16 is used to indicate that a device is unsynchronized.
Stratum 0 devices are also known as reference clocks, they are high-precision timekeeping devices such as atomic (cesium, rubidium) clocks, GPS clocks or other radio clocks. Stratum 1 are computers whose system clocks are synchronized to within a few microseconds of their attached stratum 0 devices. Stratum 1 servers may peer with other stratum 1 servers for sanity checking and backup. They are also referred to as primary time servers.
See http://en.wikipedia.org/wiki/Network_Time_Protocol#Clock_strata for more informations.
Seconds since last received packet. Should ranged from 0 to current poll value.
Value calculated by this formula <http://en.wikipedia.org/wiki/Network_Time_Protocol# Clock_synchronization_algorithm.>. Present how symmetric the round-trip time is. The more symmetric (close to 0) it is, the more accurate the estimate of the current time. Value is less than 0 means time to send packet from this machine to index server is less than time to send packet from that server back to this machine and vice versa.
clock frequency offset (PPM parts-per-million) calculated by running index daemon. This mean how different between this index to remote reference server. The closer to 0, the more precise the time is.
The estimated offset/error all the way to the stratum 1 source.
Represents the maximum error of the local clock relative to the reference clock in seconds.