Page History

...

Looking at the hourly ping losses (there are ~20 pings in an hour, so a loss of 1 ping is 5% loss) seen from SLAC for January 30th 2008 for large increases in losses which persisted to the end of the day (to avoid regular diurnal change, but unfortunately missing cases where the effect was removed by the end of the day, e.g. by re-routing), the main effects seen are shown in Table 1.  In Table 1, the Loss before is the average loss before the outage, the Loss after is the loss after the outage started. The Sites affected is the number of sites monitored in the country that observed an effect, the total is the total number of sites monitored in the country.  It is interesting that in many cases not all hosts were affected. This may be due to use of different carriers. The impact of such losses can make many applications unusable.

Table 1: Hourly Ping Losses

Since the effect may have been transitory while the data was re-routed we looked for increases in losses on January 30th.  The effect was seen in about 15 countries of the over 150 countries monitored by PingER. This is shown in Table 2 below.

Table 2: Daily Ping Losses

...

 Countries |  Throughput before (kbits/s)
|  Throughput after (kbits/s)
|  Sites affected / total
|

The above table shows the average of throughput before and after the start of the fibre outage. The results clearly show order of magnitude reductions in throughput.

...