On April 22, 2009 @ 6:00 AM PDT our network engineers noticed a drop within our connectivity to the outside world. The dropped was alerted via external
monitoring system where we quickly identified the issue and try isolating the issue as quickly as possible. Within a few minutes onsites our engineers
noticed our link lights were down which we quickly notified building engineers of the physical connection issue. Once the building engineers were onsite
they noticed a cut across multiple fibers. It appears the conduit that housed multiple fibers were cut. Many providers that carried physical transit through
the same conduit were affected. Within 1 hour our building engineers were able to spliced the fiber and bring everything online. The fiber that connects to another
secondary router that housed multiple carriers was unreachable during this time. Although we have provider redundancy our physical fiber transit were not redundant.
As of now we are in discussions with our building engineers to use other conduits to avoid issues like this again. This will allow us to achieve higher
uptimes and avoid any physical issues that may occur within our datacenter.
We are also in talk with our China providers to see if we can utilize a direct transit to China network. This will allow better latency and fewer drops
during transit. We'll update you once we make a decision vendors that will allow better transits within the Asia market.作者: cuxian 时间: 2010-4-23 12:31
等于没说什么,这些情况大家都知道了作者: fd3344520 时间: 2010-4-24 11:19
提示: 作者被禁止或删除 内容自动屏蔽