As mentioned in Case TS004024647 there are some situations where inserts in lob structures do not provide the expected performance. Instead they cause massive CPU consumption in the Coupling Facility for the spacemap search.
The problem is excessive cpu for looking up spacemap pages and reloading and swapping them in and out again repeatedly. In our case this apeared at about 45 LOB datasets (and also around 20 million rows)
|Who would benefit from this IDEA?||All customer using Lobs|
How should it work?
The insert performance should be independend from the number of datasets behind the Lob Tablespace. In the scenario of Case TS004024647 there is a situation discribed which doesn't scale as expected.
|Priority Justification||Because in Case TS004024647 thousands of customers where affected for many hours|
|Customer Name||DATEV eG|
NOTICE TO EU RESIDENTS: per EU Data Protection Policy, if you wish to remove your personal information from the IBM ideas portal, please login to the ideas portal using your previously registered information then change your email to "firstname.lastname@example.org" and first name to "anonymous" and last name to "anonymous". This will ensure that IBM will not send any emails to you about all idea submissions