DB2 Version 9.7 for Linux, UNIX, and Windows

Global health snapshots

On a partitioned database system you can take a health snapshot of the current partition, a specified partition, or all partitions. When taking a global health snapshot across all the partitions of a partitioned database, data is aggregated, where possible, before the results are returned.

The aggregated alert state for the health indicator is equivalent to the highest severity alert state across all the database partitions. Additional information and history data cannot be aggregated across the database partitions, and therefore are not available. The remaining data for the health indicator is aggregated as detailed in the table below.

Table 1. Aggregation of health indicator value, timestamp, and formula data
Health indicator	Aggregation details
db2.db2_op_status db2.sort_privmem_util db2.mon_heap_util db.db_op_status db.sort_shrmem_util db.spilled_sorts db.log_util db.log_fs_util db.locklist_util db.apps_waiting_locks db.db_heap_util db.db_backup_req ts.ts_util	The health indicator value is obtained from the partition that contains the highest value. The evaluation timestamp and formula are obtained from the same partition.
db.max_sort_shrmem_util db.pkgcache_hitratio db.catcache_hitratio db.shrworkspace_hitratio	The health indicator value is obtained from the partition that contains the lowest value. The evaluation timestamp and formula are obtained from the same partition.
db.deadlock_rate db.lock_escal_rate	The health indicator value is the sum of the values across all the database partitions. The evaluation timestamp and formula cannot be aggregated and are not available.
ts.ts_op_status tsc.tscont_op_status tsc.tscont_util	These health indicators is not aggregated.
db.hadr_op_status db.hadr_log_delay	These health indicators are not supported in a multiple partition database.
db.tb_reorg_req db.tb_runstats_req db.fed_nicknames_op_status db.fed_servers_op_status	This health indicator is evaluated only on one partition, so no aggregation is required. The data is returned from the partition which is evaluating the health indicator.

Note: When taking a global snapshot on a single partition object, the output includes all the attributes because there are no partitions to aggregate.