.ca, slowness, architecture changes, and the theory of existence

From: Charles Daminato (chuck@tucows.com)
Date: Thu Aug 22 2002 - 11:35:10 EDT


So... bouncing something off you guys:

CIRA has been working hard on developing a new architecture for their
registry systems which will GREATLY improve response times and consistency.
This system is currently active, pointing to LIVE data, but not completely
in production (they haven't made the full cutover yet).

They have asked registrars to use the system to test it out and see how it
works for them (functionally it's identical, this would simply be a server
change for us - no code would be different).

Our question is this: There is no way of knowing how their system will
handle the influx of new traffic (we'll be monitoring this). As it stands,
CIRA queries through our system (lookups, new domains, renewals,
modifications) are quite slow (and it's been this way for a LONG time). We
have the opportunity to test out the new system with real-time traffic, the
only caveat is that this isn't a full production system yet (i.e. CIRA
hasn't done a full systems change, which would be invisible, so it could -
in some ways - be considered beta).

The benefit is we could experience and extreme speed increase, the risk is
that their system may not handle the influx of traffic - we'd hit a bump and
have to go back to the current system.

How do you feel about this?

As a side note: If Schrodinger and Heisenberg went into a black box to
discuss the matter of the chicken vs. the egg, would the discussion actually
happen?

Charles Daminato
OpenSRS Product Manager
Tucows Inc. - chuck@tucows.com



This archive was generated by hypermail 2.1.3 : Tue Oct 19 2004 - 23:37:24 EDT