uptime graph, HAL9000, 24 hour view:
uptime graph, HAL9000, 96 hour (4 day) view:
A note Nov-2001
Verizon have been fairly good at keeping HAL9000's connection up. As was noted in a previous news note in this space, I anticipated that Telocity (now DirecTV DSL) might issue a disconnect order for my "data" phone line (to which I had Verizon connect ADSL). Well, thankfully, Verizon did not allow Telocity (or anyone else for that matter) to stop DSL from working on my line, so everything hummed along, virtually without interruption.
BroadbandReports.com (BBR) monitors hosts by sending out a few Internet Control Message Protocol (ICMP) echo request ("ping") packets every ten minutes. It measures how long it took to receive an ICMP echo reply (round trip time, or "RTT"), and how many replies never got back to the monitoring host (packet loss).
Time increases along the X axis on these graphs. RTT is plotted in green along the Y axis. Packet loss percent is plotted in blue, also on the Y axis. Obviously, the lower everything is plotted on the graph, the better. If the packet loss figure ever rises above about 5%, maintaining a connection with HAL9000 will be difficult (IP is generally designed to tolerate 2% or less packet loss). If packet loss ever reaches 100%, that means HAL9000 is totally offline. If the graph ever shows packet loss above 30%, it probably means that HAL9000's DSL modem was resynchronizing often. (One or two resyncs would not show up on the graph because the data are averaged together, and it's only sampled every 10 minutes.)
If there is ever a blank space in the graph, it is usually for one of two reasons: