Hi!
I've noticed a strange behavior of HANA (SPS 08 and 09) with tables loading in memory. I have a big column table partitioned by HASH on the first level and by RANGE (calendar period) on the second level. For clean test I unload the table comletely from memory. After that I run an SQL query with the restriction of calendar period in the way that only single or very few partitions must be analyzed. If that query is missing in SQL plan cache HANA loads all partitions in memory, but only the columns involved in the query. After that I unload the table from memory again and repeat the query. At this moment a query is already in SQL plan cache and only relevant partitions are loaded. This behavior is normal for HANA? The most recent revisions work the same way? If it's normal, please, provide me to official documentation which desrcibe this.
Thanks, Andrey.