1. UNIFYBroker Forum
  2. Bugs
0
Fixed

High memory consumption when change detection affects large datasets

Matthew Clark 12 years ago • updated by anonymous 8 years ago • 4

QDET-194 shows that if a relational change affects a large data set, memory consumption can grow quite high and result in OutOfMemory exceptions. Changes should be made to the transformation contributions such that they do not load all potential matches immediately into memory.

Matthew Clark 12 years ago

A potential fix has been pushed to the v3.0.7 branch for the RelationalTransformationContribution. This will be uploaded to QDET-194 for confirmation of changes.

For v4.0, these fixes will need to be pulled over and added to the other relational contribution types as well, such as the multi source relational contribution.

Matthew Clark 12 years ago

Adam,

Following feedback on QDET-194, update the other transformation contribution types to use this paging algorithm.

Adam van Vliet 12 years ago

Migrated fixes from v3.0. Fixed up other contributions.

Please confirm.

Thanks.

Matthew Clark 12 years ago

Changes are present in v4.0, but we lack sufficient data to completely replicate and directly confirm this for v4.0. Issue closed.

Customer support service by UserEcho