Eskimo North


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Dial Outage Resolve shortly after 8pm




     The problems with our POPs have been resolved shortly after 8pm.

     Technical explanation follows:

     We lease ports on RAS's at ELI on a combination of Cisco 5300's and
5800's.  Normally, the IP space assigned to a dynamic IP customer via DHCP
would be from a pool internal to the Cisco.

     But ELI leases ports to many different customers and has POPs in many
locations.  Maintaining the configuration for each RAS internally would be
a major headache.

     So instead they have a system that serves as an IP pool allocator they
call the master IP pool allocator.  Basically, each customers IP space and
ports are programmed into that box. When a call hits a port, the Cisco queries
that system for the IP to assign to it.  The system gives the Cisco RAS an IP
address for the customer in question, and then the Cisco assigns that to the
customer via DHCP.  That way they can administer the IP space for many
customers and many RAS's from one centralized system.

     It is that system that failed today.  They tell me this is the first time
that it has ever failed so we should not expect it to be a frequent source of
trouble.  It is the first time it has failed as long as we've been leasing
ports there which has been more than eight months.

     Total downtime appears to be about an hour and fifteen minutes or so, this
failed around 6:45pm was restored at approximately 8pm.