Table and Index Positions.
Posted in 2016
Topics: Third-Party Tools & Monitoring
In OAT / Profiled Fields / ISAM, there is an entry called "TABLE AND INDEX PSOSITONS". Can anyone tell me what this actually means? I have an 11.50 sever that runs perfectly well for about 90 days and then users report ever increasing slowness. A reboot solves the problem. I suspect memory. Investigation shows Memory Sorts take a jump in baseline of about 0 to about 2000 and at the same time Table and index positions start to increase linearly and quite steeply. Selected rows also increases linearly. 3 days after this event, used virtual memory also jumps significantly. However there is no change in the user usage pattern that could account for the increases. Ray
Position is how many times the engine has to position the scan at a specific value in the index. If you want to locate a specific single unique value in an index, then a position would be 1 If you want to scan a specific range of values, ie from "fred" to "Gerry" then a position would be 1. If you want to look a several unique values, unique serial id 135 and 147 and 245, then position would be 3 If you want to scan a several ranges, then positions would be 1 for each range. Hop this helps, John F. Miller III miller3@us.ibm.com ids-bounces@iiug.org wrote on 10/10/2016 12:21:46 PM: > From: "RAY BURNS" <ray.burns@velocityglobal.co.nz> > To: ids@iiug.org > Date: 10/10/2016 12:22 PM > Subject: Table and Index Positions. [37959] > Sent by: ids-bounces@iiug.org > > In OAT / Profiled Fields / ISAM, there is an entry called "TABLE AND INDEX > PSOSITONS". Can anyone tell me what this actually means? > > I have an 11.50 sever that runs perfectly well for about 90 days and then > users report ever increasing slowness. A reboot solves the problem. I suspect > memory. Investigation shows Memory Sorts take a jump in baseline of > about 0 to > about 2000 and at the same time Table and index positions start to increase > linearly and quite steeply. Selected rows also increases linearly. 3 days > after this event, used virtual memory also jumps significantly. > However there is no change in the user usage pattern that could account for > the increases. > > Ray > > > ***************************************************************************= **** > Forum Note: Use "Reply" to post a response in the discussion forum. >