> From January 5-12, we had an outage for cooling pipeline > upgrade, and we agreed with SCCS to close batch nodes > from midnight, January 3. That is a period of 9 ~ 10 > days. I think that is the ~ 2/3 availability data coming > from. There is another discussion going on where "scheduled outages" for a site will be accounted for in the next (or next next?) revision of the availability reporting to address the confusions/impressions that can occur with the existing single viewpoint numbers. Gary