Index Activity (i)


This panel shows detailed index activity for all indexes being monitored, based on your table and index range parameter settings. By default, only the first 50 application tables and indexes are monitored. See Table and Index Range Information (T) for more information and suggested values.

ProTop RT Label Alertable Metric* Description
Idx# idxNum Unique index reference from Progress schema table _index
Area# idxAreaNum Database area containing the index
BX idxPool Blocks assigned to the primary buffer pool (B1) or alternate buffer pool (B2)
Index Name idxName Name of the index
Blocks idxBlks Size of the index in blocks
Util idxUtil Percent utilization of block space by the index
Lvls idxLvls Number of levels from index root to leaf nodes
Idx Root idxRoot Block ID of the index root
Note idxNote Flags on the index (P)rimary, (U)nique, (W)ord, and (i)nactive
Create idxCr Number of index entries created
Read idxRd Number of index reads
Split idxSp Number of index block splits
Delete idxDl Number of index entries deleted
BlkDl idxBlkDl Number of index blocks deleted

*Requires data collector "IndexActivity" in pt3agent.cfg. With the commercial version of ProTop, you can configure alerts using the metric names in this column. For a full list of metrics, see the Alertable Metrics section.

Insight into Index Activity

Percent Utilization: Indexes blocks holding indexes of actively modified tables will invariably become more sparse, and as they become more sparse, it will take more disk IO and more buffer cache memory to access them. Essentially, you are reading a whole bunch of empty space!

Once percent utilization falls below 70%, you should consider compacting those indexes. It isn't necessarily urgent but you should add it to your task list. Search through our blog page for discussions on index compact.

One note about index compacting: on actively modified tables, you do not want to compact to 100% as the next add or update could cause an expensive block split. Instead, compact to 90%.