From issues-return-193244-archive-asf-public=cust-asf.ponee.io@spark.apache.org Fri Jun 1 22:03:04 2018 Return-Path: X-Original-To: archive-asf-public@cust-asf.ponee.io Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by mx-eu-01.ponee.io (Postfix) with SMTP id 873E618063A for ; Fri, 1 Jun 2018 22:03:03 +0200 (CEST) Received: (qmail 26531 invoked by uid 500); 1 Jun 2018 20:03:02 -0000 Mailing-List: contact issues-help@spark.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Delivered-To: mailing list issues@spark.apache.org Received: (qmail 26522 invoked by uid 99); 1 Jun 2018 20:03:02 -0000 Received: from pnap-us-west-generic-nat.apache.org (HELO spamd4-us-west.apache.org) (209.188.14.142) by apache.org (qpsmtpd/0.29) with ESMTP; Fri, 01 Jun 2018 20:03:02 +0000 Received: from localhost (localhost [127.0.0.1]) by spamd4-us-west.apache.org (ASF Mail Server at spamd4-us-west.apache.org) with ESMTP id 25629C0199 for ; Fri, 1 Jun 2018 20:03:02 +0000 (UTC) X-Virus-Scanned: Debian amavisd-new at spamd4-us-west.apache.org X-Spam-Flag: NO X-Spam-Score: -110.301 X-Spam-Level: X-Spam-Status: No, score=-110.301 tagged_above=-999 required=6.31 tests=[ENV_AND_HDR_SPF_MATCH=-0.5, RCVD_IN_DNSWL_MED=-2.3, SPF_PASS=-0.001, USER_IN_DEF_SPF_WL=-7.5, USER_IN_WHITELIST=-100] autolearn=disabled Received: from mx1-lw-us.apache.org ([10.40.0.8]) by localhost (spamd4-us-west.apache.org [10.40.0.11]) (amavisd-new, port 10024) with ESMTP id GVVVZJA8bwAS for ; Fri, 1 Jun 2018 20:03:01 +0000 (UTC) Received: from mailrelay1-us-west.apache.org (mailrelay1-us-west.apache.org [209.188.14.139]) by mx1-lw-us.apache.org (ASF Mail Server at mx1-lw-us.apache.org) with ESMTP id 23BEB5F57D for ; Fri, 1 Jun 2018 20:03:01 +0000 (UTC) Received: from jira-lw-us.apache.org (unknown [207.244.88.139]) by mailrelay1-us-west.apache.org (ASF Mail Server at mailrelay1-us-west.apache.org) with ESMTP id A711CE012B for ; Fri, 1 Jun 2018 20:03:00 +0000 (UTC) Received: from jira-lw-us.apache.org (localhost [127.0.0.1]) by jira-lw-us.apache.org (ASF Mail Server at jira-lw-us.apache.org) with ESMTP id 6891421094 for ; Fri, 1 Jun 2018 20:03:00 +0000 (UTC) Date: Fri, 1 Jun 2018 20:03:00 +0000 (UTC) From: "Reynold Xin (JIRA)" To: issues@spark.apache.org Message-ID: In-Reply-To: References: Subject: [jira] [Commented] (SPARK-24374) SPIP: Support Barrier Scheduling in Apache Spark MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable X-JIRA-FingerPrint: 30527f35849b9dde25b450d4833f0394 [ https://issues.apache.org/jira/browse/SPARK-24374?page=3Dcom.atlassia= n.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D164= 98493#comment-16498493 ]=20 Reynold Xin commented on SPARK-24374: ------------------------------------- Just thought of this =E2=80=94 Continuous Processing really requires gang s= cheduling. As it is today, I think things just hang if not all tasks are sc= heduled. =C2=A0 > SPIP: Support Barrier Scheduling in Apache Spark > ------------------------------------------------ > > Key: SPARK-24374 > URL: https://issues.apache.org/jira/browse/SPARK-24374 > Project: Spark > Issue Type: Epic > Components: ML, Spark Core > Affects Versions: 3.0.0 > Reporter: Xiangrui Meng > Assignee: Xiangrui Meng > Priority: Major > Labels: SPIP > Attachments: SPIP_ Support Barrier Scheduling in Apache Spark.pdf > > > (See details in the linked/attached SPIP doc.) > {quote} > The proposal here is to add a new scheduling model to Apache Spark so use= rs can properly embed distributed DL training as a Spark stage to simplify = the distributed training workflow. For example, Horovod uses MPI to impleme= nt all-reduce to accelerate distributed TensorFlow training. The computatio= n model is different from MapReduce used by Spark. In Spark, a task in a st= age doesn=E2=80=99t depend on any other tasks in the same stage, and hence = it can be scheduled independently. In MPI, all workers start at the same ti= me and pass messages around. To embed this workload in Spark, we need to in= troduce a new scheduling model, tentatively named =E2=80=9Cbarrier scheduli= ng=E2=80=9D, which launches tasks at the same time and provides users enoug= h information and tooling to embed distributed DL training. Spark can also = provide an extra layer of fault tolerance in case some tasks failed in the = middle, where Spark would abort all tasks and restart the stage. > {quote} -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org For additional commands, e-mail: issues-help@spark.apache.org