Greg's network link statistics

This page shows my network availability and performance statistics connected to the Australian National Broadband Network. It bases on earlier pages that I wrote for the DSL and satellite connections.

I had hoped that with the advent of the NBN, this page would no longer be needed. Unfortunately, this has proven not to be the case. While I don't have issues with throughput, the availability has been the worst I have ever experienced, with outages of up to 48 hours. It's not clear where the problem lies, but I have observed serious issues with the DHCP authentication that my ISP uses. More here when I understand the situation better.

The graphs show:

Network status. This is the number of remote systems responding to a ICMP echo (ping) packet at the same time as the link status ping. I currently ping 5 systems which I frequently access: freefall.freebsd.org (now somewhere in New York state), www.lemis.com, my external web server in Raleigh NC, ffm.lemis.com, located in Frankfurt am Main, ozlabs.org, located in Canberra, and ftp.netbsd.org, located somewhere in the USA. Normally this value should be 5. If it's less, the script retries every second until full connectivity is restored.
Link packet loss, which has proven to be a significant issue since early 2018. I send 20 ICMP ping messages to the other end of the network link and count how many are lost.

I previously had parameters "link status" and "TCP speed”, but they no longer seem to make much sense.

The graphs were made with gnuplot, and I'm not very happy with the smoothing. In particular, the right-hand side of smoothed graphs has too much influence, and it's easy to get the false impression that the link status or TCP speed have changed significantly in the last few minutes. If anybody can point me at a way to fix this problem, I'd be grateful.

Click on the graphs for a 1600x1200 version.

Detailed outage information

This section monitors the NBN link for outages, including planned outages. I have manually edited the data to remove false positives, such as when moving house. I have a program that evaluates the outage information and produces detailed outage information. Here's the information on the past 20 outages, along with the overall statistics:

Start time End time  Duration   Badness        from                    to
                     (seconds)

1583380701 1583380725     24	 29.032	#  5 March 2020 14:58:21  5 March 2020 14:58:45
1586970212 1586971656   1444	  0.001	# 16 April 2020 03:03:32 16 April 2020 03:27:36
1588248517 1588251543   3026	  0.003	# 30 April 2020 22:08:37 30 April 2020 22:59:03
1589207590 1589208307    717	  0.004	# 12 May 2020 00:33:10 12 May 2020 00:45:07
1589728447 1589730531   2084	  0.007	# 18 May 2020 01:14:07 18 May 2020 01:48:51
1594289411 1594289414      3	  0.001	#  9 July 2020 20:10:11  9 July 2020 20:10:14
1594736249 1594736319     70	  0.008	# 15 July 2020 00:17:29 15 July 2020 00:18:39
1594833568 1594833638     70	  0.037	# 16 July 2020 03:19:28 16 July 2020 03:20:38
1596474098 1596475481   1383	  0.002	#  4 August 2020 03:01:38  4 August 2020 03:24:41
1602175169 1602175250     81	  0.001	#  9 October 2020 03:39:29  9 October 2020 03:40:50
1605191195 1605191319    124	  0.001	# 13 November 2020 01:26:35 13 November 2020 01:28:39
1605196381 1605196463     82	  0.711	# 13 November 2020 02:53:01 13 November 2020 02:54:23
1605798510 1605798735    225	  0.006	# 20 November 2020 02:08:30 20 November 2020 02:12:15
1606140598 1606140735    137	  0.011	# 24 November 2020 01:09:58 24 November 2020 01:12:15
1606162445 1606162502     57	  0.166	# 24 November 2020 07:14:05 24 November 2020 07:15:02
1606162917 1606162967     50	  8.675	# 24 November 2020 07:21:57 24 November 2020 07:22:47

Summary
Total 453 outages, total time 1049851 seconds (12 days, 03:37:31)
Longest outage:			192150 seconds (2 days, 05:22:30)
   Start:			21 February 2015 04:36:54
   End:				23 February 2015 09:59:24
Average time between outages:	472212 seconds (5 days, 11:10:12)
Average duration:		2317 seconds (00:38:37)
Availability:			99.51%

“Badness” is an attempt to quantify the effect. It's the reciprocal of the number of seconds per hour that the link was up between failures (i.e. 3600 / uptime).

And here is the summary information for the past 10 days with a less than 100% record:

Timestamp      Outages  Duration  Availability    Date
                        (seconds)
1589205600	  1	   717	 99.17%	# 12 May 2020
1589724000	  1	  2084	 97.59%	# 18 May 2020
1594216800	  1	     3	100.00%	#  9 July 2020
1594735200	  1	    70	 99.92%	# 15 July 2020
1594821600	  1	    70	 99.92%	# 16 July 2020
1596463200	  1	  1383	 98.40%	#  4 August 2020
1602162000	  1	    81	 99.91%	#  9 October 2020
1605186000	  2	   206	 99.76%	# 13 November 2020
1605790800	  1	   225	 99.74%	# 20 November 2020
1606136400	  3	   244	 99.45%	# 24 November 2020

The dates in the left columns are in UNIX time_t format to ease further processing. I also have lists of all individual outages, or summaries per day.