TB> it seems ironic that projects of such importance to the IP World can be TB> configured to not to withstand a single platform crash Jake replies:
That statement is untrue. At all our production sites, the service can withstand a single platform crash. This is because we have redundancy at those exchange points. However, at MAE-West, there was only one machine operational. The lack of a redundant machine was what prohibited declaration of production status.
Jake, I am not passing judgment here nor trying to be antagonistic (I do however, *attempt* to use independent good judgment). It just appears to this chair, from observation of the situation, there were some IP customers peering with the MAE-WEST RA, and that when the server died, their was a panic to get it back on line because it effected their customers. Maybe this perception is incorrect and no one was effected or experienced a loss of service; in that event, agreed. But on the other hand, even though 'technically or administratively' MAE_WEST RA was of named status 'not-production'; when it crashed if it effected the daily production operation of networks, then questions and introspection are expected. Polemically speaking (as Aristotle might say), one could make some strong points that MAE-WEST was being 'productive' ;-) De facto vs. de jure... Again, I am not passing judgment, it just appears to me that on a $10,000,000 dollar program at an important exchange such as MAE-WEST (whatever the contractor chooses to call the status 'officially'); a little redundancy and prudent engineering and benefits the community. Best Regards, Tim ---