[HBASE-13109] Make better SEEK vs SKIP decisions during scanning

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.0.1, 1.1.0, 0.98.12, 2.0.0
• Hadoop Flags: Reviewed
• Assignee: Lars Hofhansl
• Reporter: Lars Hofhansl
• Created: 26/Feb/15 02:34
• Updated: 18/Oct/17 09:46
• Resolved: 17/Mar/15 16:07

Description

I'm re-purposing this issue to add a heuristic as to when to SEEK and when to SKIP Cells. This has come up in various issues, and I think I have a way to finally fix this now. HBASE-9778, HBASE-12311, and friends are related.

— Old description —

This is a continuation of HBASE-9778.
We've seen a scenario of a very slow scan over a region using a timerange that happens to fall after the ts of any Cell in the region.
Turns out we spend a lot of time seeking.

Tested with a 5 column table, and the scan is 5x faster when the timerange falls before all Cells' ts.
We can use the lookahead hint introduced in HBASE-9778 to do opportunistic SKIPing before we actually seek.

Issue Links

• is related to: HBASE-9778 Add hint to ExplicitColumnTracker to avoid seeking
• is related to: PHOENIX-1731 Add getNextIndexedKey() to IndexHalfStoreFileReader and FilteredKeyValueScanner
• is related to: HBASE-9769 Improve performance of a Scanner with explicit column list when rows are small/medium size
• relates to: HBASE-19034 Implement "optimize SEEK to SKIP" in storefile scanner
• relates to: HBASE-17958 Avoid passing unexpected cell to ScanQueryMatcher when optimize SEEK to SKIP
• relates to: PHOENIX-1731 Add getNextIndexedKey() to IndexHalfStoreFileReader and FilteredKeyValueScanner