Unlike sitting in gridlock on the M1 and having no confirmed clue about the cause, one can use traceroute on the Internet to get a fairly good idea of where what is broken. Then one can mosey over to that network, from another continent with access (or better access) and see when things will be running at top speed again.
Recently I had cause to move around 22TB of data from Europe and that is when one feels even small speedbumps on the road.
Anyway on this sunny, warm and relaxed Friday afternoon the IS network International connectivity is up and down. HTTP traffic was not designed with route failures in mind, unlike smtp and other protocols that can survive moments or even hours of outages, http is very unforgiving… Anyway, so I hop over to Vancouver in Canada to have a look at South Africa from the outside. The shortest route it seems is to Washington and then to NY in the USA. From there London, UK and then SA, all in 334 odd ms, when all is working well. Surprise surprise, when all is not working well it dies locally. Interesting that is…
Traceroute from CA to SA:
bx2-seattle_GE3_0_0.net.bell.ca 4.402 ms
ae-7.r21.sttlwa01.us.bb.gin.ntt.net 4.426 ms
ae-3.r22.nycmny01.us.bb.gin.ntt.net 103.712 ms
as-1.r22.londen03.uk.bb.gin.ntt.net 156.733 ms 139.372 ms
ae-1.r02.londen03.uk.bb.gin.ntt.net 157.207 ms
dimensiondata-0.r02.londen03.uk.bb.gin.ntt.net 165.779 ms
cdsl2-rba-te1-1.ip.isnet.net 334.732 ms