Return-Path: X-Original-To: apmail-accumulo-notifications-archive@minotaur.apache.org Delivered-To: apmail-accumulo-notifications-archive@minotaur.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id 72B8ACE63 for ; Wed, 17 Dec 2014 19:01:14 +0000 (UTC) Received: (qmail 18917 invoked by uid 500); 17 Dec 2014 19:01:14 -0000 Delivered-To: apmail-accumulo-notifications-archive@accumulo.apache.org Received: (qmail 18883 invoked by uid 500); 17 Dec 2014 19:01:14 -0000 Mailing-List: contact notifications-help@accumulo.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: jira@apache.org Delivered-To: mailing list notifications@accumulo.apache.org Received: (qmail 18854 invoked by uid 99); 17 Dec 2014 19:01:14 -0000 Received: from arcas.apache.org (HELO arcas.apache.org) (140.211.11.28) by apache.org (qpsmtpd/0.29) with ESMTP; Wed, 17 Dec 2014 19:01:14 +0000 Date: Wed, 17 Dec 2014 19:01:14 +0000 (UTC) From: "Christopher Tubbs (JIRA)" To: notifications@accumulo.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Commented] (ACCUMULO-96) Need to handle case where tablet has lots of delete entries MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 [ https://issues.apache.org/jira/browse/ACCUMULO-96?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14250319#comment-14250319 ] Christopher Tubbs commented on ACCUMULO-96: ------------------------------------------- I think the case still has merit. I don't know if anybody's interested in working on it. > Need to handle case where tablet has lots of delete entries > ----------------------------------------------------------- > > Key: ACCUMULO-96 > URL: https://issues.apache.org/jira/browse/ACCUMULO-96 > Project: Accumulo > Issue Type: Bug > Components: tserver > Reporter: Keith Turner > > Currently Accumulo only drops delete entries when its major compacting all files. All files may never be compacted. Lots of delete entries building up can be problematic for scan performance. > Possible solutions : > * Keep a count of the # of delete entries per file. Compact all when the ratio of delete entries to entries crosses a threshold. > * When a scan encounters a tablet w/ a lot of delete entries, schedule a majc. This is reactive, but removes the need for bookkeeping in the previous solution. -- This message was sent by Atlassian JIRA (v6.3.4#6332)