It has come to our attention that at approximately 2:00pm Sydney time, many of our customers began to experience issues relating to internet connectivity and performance.
We have received information that a core router within the Telstra internet backbone has failed which has caused routing issues and delays. This information was sourced from Whirlpool.net.au (http://forums.whirlpool.net.au/forum-replies.cfm?t=1871831&p=10) and confirmed by a Telstra contact.
We are in the process of obtaining more information and will let our customers know as information becomes available.
You may find that some web sites, email transport and banking/atm services will be affected.
Update: 15:00
Fault appears to have been with core routers within the Telstra network. Apparently, a lesser ISP had sent their entire BGP table to Telstra causing the router(s) to fail.
This appears to have been resolved and internet should be starting to come back online for everyone.
For information regarding how BGP works, please see:
http://en.wikipedia.org/wiki/Border_Gateway_Protocol
For a list of all networks affected by this issue, please see:
http://bgp.he.net/AS1221#_peers
Update 17:00
Confirmed that all issues have been resolved, and the actual cause is as follows:
At 13:40 BGP peering shut down on all Reach International gateways, due to the maximum prefix limit being exceeded. The sudden very latge increase in advertisted routes that triggered this has been attributed to Dodo. Dodo stopped advertising the routes and BPG came up, restoring all international service at approv 14:30. As mitigation, IP Core are implementing a change today to Dodo’s configuration to prevent any further flooding of advertised prefixes from this peer.
News articles are now appearing reflecting this information: