Track index bloat estimate
What does this MR do?
This adds tracking of the bloat estimate for a postgres index while reindexing. The bloat estimate is used to pick the most impactful indexes for reindexing. We start to track what the bloat estimate was here. This allows us to check the quality of the estimate later: When reindexing has finished, the difference in index size should roughly match the bloat estimate.
This is merely a way for us to check this and learn more about the estimation quality, there is no more use to this as of now.
Does this MR meet the acceptance criteria?
Conformity
Edited by Andreas Brandl