[geomesa-dev] [JIRA] (GEOMESA-2308) Command line tool for incrementally

["Emilio Lahr-Vivaz (JIRA)" <jira@xxxxxxxxxxxxxxxxxxxxx>]

Title: Message Title

Emilio Lahr-Vivaz created an issue

GeoMesa / GEOMESA-2308

Command line tool for incrementally compacting accumulo tables

Issue Type:	Improvement
Assignee:	Unassigned
Created:	20/Jun/18 12:19 PM
Priority:	Major
Reporter:	Emilio Lahr-Vivaz

In order to ensure aged-off data is deleted from disk, we need to compact the tables.

The compact command takes multiple ranges. That way you can optimize at least some of the compaction and not trigger the entire table at once.

For Z3 you can avoid date ranges that are newer.

For the other tables (e.g. record and attributes) that aren't using z3 uuid or z3 attr. You can list all the splits and then walk through them in some small incremental process...this will avoid the queuing of large numbers of compactions. However, you can also just decrease the number of total compactions allowed and then queue up the entire table...But the dangerous part of this is that the table can reach file.max and then I don't know if it will wait until major compaction finishes before allowing minc. Doing the tiny boring incremental compaction fixes that.

Add Comment

Get Jira notifications on your phone! Download the Jira Cloud app for Android or iOS
This message was sent by Atlassian Jira (v1001.0.0-SNAPSHOT#100087-sha1:6ad1004)