Database Reference
In-Depth Information
Step 3
With the priority set to LOW, the tables identified in Step 1 will be auto-populated into the in-memory area of the
SGA. This can be validated using the following script:
Script: MVRACPDnTap_impopcheck.sql
SELECT ims.inst_id INT,
ims.owner OWNER,
ims.segment_name SEGNAME,
ims.segment_type SEGTYPE,
ims.populate_status STATUS,
ims.inmemory_priority PRIORITY,
ims.inmemory_distribute DISTRIBUTE,
ims.inmemory_duplicate DUPLICATE,
ims.inmemory_compression COMPRESSD,
ims.bytes /ims.inmemory_size comp_ratio
FROM gv$im_segments ims;
In the above output, the following are the column descriptions:
OWNER is the owner of the object in the list.
SEGNAME is the segment name/object name loaded into the in-memory area.
SEGTYPE is the type of segment loaded into the in-memory area, such as a table, index
partition, etc.
STATUS is the current status of the operation. Valid values are STARTED and COMPLETED .
PRIORITY specifies the priority to use when populating tables in the IM column store.
DISTRIBUTE indicates how the data is distributed among the RAC instances in the cluster.
DUPLICATE specifies how many copies of each In-Memory Compression Unit (IMCU) of the
tables in the IM column store will be spread across all the Oracle RAC instances.
COMPRESSED specifies that in-memory compression should be used for the instance.
COMP RATIO is the current ratio of compression.
Most of the data are attributes or settings are assigned to the object through an ALTER statement.
The speed of data population depends on the number of in-memory server processes configured, which is
controlled using the parameter INMEMORY_MAX_POPULATE_SERVERS . The server processes are identified by
ora_wnnn_dbname , for example
SQL> !ps -ef | grep ora_w
oracle 11913 1 0 15:17 ? 00:00:00 ora_w001_SSKYDB_2
oracle 14199 1 0 15:22 ? 00:00:09 ora_w002_SSKYDB_2
oracle 31054 1 0 16:07 ? 00:00:00 ora_w003_SSKYDB_2
 
Search WWH ::




Custom Search