Today, I'm introducing a new metric, one that I've been working on and refining for some time: The Grid Stability Index (Linden Lab's release of the first of a series of quality-of-service metrics today is entirely coincidental). It's an open-ended scale, starting at zero (stable) and increasing as indicators show variance from a stable situation (including both planned and unplanned outages, service interruptions, login failures, and widespread performance problems).
Yesterday, for example, with downtime and region issues came in at 158.65 [New Scale 22.03]. Complete downtime for a 24 hour period would rate 280, but - considering that amounts to a comparatively stable (if undesirable) situation - there are many issues which could potentially cause the number to exceed that value (the highest recorded is 1,365.94 one year ago on the 19th of September 2006).
In a sense, you can also think of this as a grid instability index, if you prefer. Either way, it gives us a metric for comparison. Just remember, lower numbers are better. We'll bring you today's figure at midnight, SLT (US Pacific).
