Return-Path: X-Original-To: apmail-hbase-user-archive@www.apache.org Delivered-To: apmail-hbase-user-archive@www.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id 893C510EE8 for ; Fri, 30 Aug 2013 20:56:44 +0000 (UTC) Received: (qmail 5745 invoked by uid 500); 30 Aug 2013 20:56:42 -0000 Delivered-To: apmail-hbase-user-archive@hbase.apache.org Received: (qmail 5687 invoked by uid 500); 30 Aug 2013 20:56:42 -0000 Mailing-List: contact user-help@hbase.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: user@hbase.apache.org Delivered-To: mailing list user@hbase.apache.org Received: (qmail 5679 invoked by uid 99); 30 Aug 2013 20:56:42 -0000 Received: from nike.apache.org (HELO nike.apache.org) (192.87.106.230) by apache.org (qpsmtpd/0.29) with ESMTP; Fri, 30 Aug 2013 20:56:42 +0000 X-ASF-Spam-Status: No, hits=2.2 required=5.0 tests=HTML_MESSAGE,RCVD_IN_DNSWL_NONE,SPF_PASS X-Spam-Check-By: apache.org Received-SPF: pass (nike.apache.org: local policy) Received: from [216.39.63.145] (HELO nm27-vm5.access.bullet.mail.gq1.yahoo.com) (216.39.63.145) by apache.org (qpsmtpd/0.29) with ESMTP; Fri, 30 Aug 2013 20:56:33 +0000 Received: from [216.39.60.165] by nm27.access.bullet.mail.gq1.yahoo.com with NNFMP; 30 Aug 2013 20:56:11 -0000 Received: from [216.39.60.230] by tm1.access.bullet.mail.gq1.yahoo.com with NNFMP; 30 Aug 2013 20:56:10 -0000 Received: from [127.0.0.1] by omp1001.access.mail.gq1.yahoo.com with NNFMP; 30 Aug 2013 20:56:10 -0000 X-Yahoo-Newman-Property: ymail-3 X-Yahoo-Newman-Id: 977822.38184.bm@omp1001.access.mail.gq1.yahoo.com Received: (qmail 36231 invoked by uid 60001); 30 Aug 2013 20:56:10 -0000 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yahoo.com; s=s1024; t=1377896170; bh=RI8GayekiBrxbP+QD0lSC8oCsy/nf2SsSXSoJK4psPo=; h=X-YMail-OSG:Received:X-Rocket-MIMEInfo:X-Mailer:References:Message-ID:Date:From:Reply-To:Subject:To:In-Reply-To:MIME-Version:Content-Type; b=wPprJTQ2ljIa/lB6Le7ZRZpUMRjF9El2IYgbJkfKBbjvDP/8r0tkdheRgpeiAbfDzppY/FXGGZNPw2cxj3PzULHvthVnW/iIpfZ+TI4rRKQrl//IkZcSeg1RgiDzp5jmw6IQqiERnj8/+5rYFtEQjRVBQe/PzbwCX1cpOwpN36g= DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=X-YMail-OSG:Received:X-Rocket-MIMEInfo:X-Mailer:References:Message-ID:Date:From:Reply-To:Subject:To:In-Reply-To:MIME-Version:Content-Type; b=QjUp73gAKBQaFR+OLgt3g9KhVbxk1jFRh6mlWdHRy24Sven75ZY9K/HpICxxnC6QNWa8NKT8iqGf9Pe4ZEB7cNg701SfzciiM2LFDVwfq2Pv9+OKgqAJFUhACBqaQL5yajhur/DrlUaK77eBqi1FEjy7+kCzwC9TDPyjHk2zWQI=; X-YMail-OSG: lu.x7FQVM1n6YW8kl5nxarMk7wJSZf5er9_lwmLK1bV3DDy fPtQyhpd82SURlNLWQ5ghHNvIIEUXoQkKyxHOEhxg4ugLmfhUEQs9NZPXlYY 5PX2owZ1h4_LJbqttFlAOGjI_xxn1LuP6NlKejdEBFj1hvY_YX27NsxLuq7p .DFSsFrHsI9oGZyXBIqXMP18bjgY7HnL2hmWAysNNQzhxE2pC8mhd6aQ_cyy CZQaHxD5PlwH2tZMR0qtWiN9yM8qvYqYeJepmCMWMdA3aYKtOtgnwrh2f7Zg 07KpLsh3zcNWcl2QCGD8RCu_2PnoPE7lEgYVcDBwzX9O4DU567se9cYhh.am osTs0tT04xNq7gsb1.nuUotYhO15rNOZs3s6BHpE0XP4H1IHWa7acZg2p7Rw OgwmgdC7u7e.ZOmriqrnVoZjf4MbANziQLfntRJQpWam6_oDyrMWlE2sNFA8 zZIEIAZcAX4EJwbE.XmUi5w0pJjxKb0wMUcUMCOVrWv9EJmr3XZqULqJPw.7 OHfXoH7KKLrkyXlbaYs8V0kKrFAW4cMgSQ37sZMMXcnZdIqiKPv705.PuiYz Jyy2x9pz.m66i8nNaWKiW10_YTqSpjbWqosIZBjm_LZLp5AXEHp4ES50H044 jGwPquvHLFuuNXU7TKGHf2ssyQoEDjoyeb0Hu1aTPtO8K4WAdg_srnBP90xO 8z8bmIZ7MPL7L5ao- Received: from [12.27.66.5] by web181604.mail.ne1.yahoo.com via HTTP; Fri, 30 Aug 2013 13:56:10 PDT X-Rocket-MIMEInfo: 002.001,VmxhZCwKSSBhbSBydW5uaW5nIGluIHNlcGFyYXRlIHByb2Nlc3Nlcy4gQnV0IEkgdGhpbmsgSSBmb3VuZCBhdCBsZWFzdCBwYXJ0IG9mIHRoZSBwcm9ibGVtLgpJdCBpcyB0aGUgSFRhYmxlIGNvbm5lY3Rpb24gc2V0dXAgdGhhdCBpcyBhZmZlY3RpbmcgaXQKSWYgSSBzZXR1cCBhbGwgdGhlIGNvbm5lY3Rpb25zIGluIGFsbCB0aGUgcHJvY2Vzc2VzIGZvciBhbGwgdGhlIHRhYmxlcyB0aGVuIGFuZCB0aGVuIGNhbGwgdGhlIGVuZHBvaW50cyBpdCBnb2VzIHRocm91Z2ggb2suCkF0IGxlYXN0IGZvciB0aGUgZXgBMAEBAQE- X-Mailer: YahooMailWebService/0.8.156.576 References: <1377852561.60463.YahooMailNeo@web181606.mail.ne1.yahoo.com> Message-ID: <1377896170.4191.YahooMailNeo@web181604.mail.ne1.yahoo.com> Date: Fri, 30 Aug 2013 13:56:10 -0700 (PDT) From: Kiru Pakkirisamy Reply-To: Kiru Pakkirisamy Subject: Re: Coprocessor concurrency To: "user@hbase.apache.org" In-Reply-To: MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="2017413661-1111410278-1377896170=:4191" X-Virus-Checked: Checked by ClamAV on apache.org --2017413661-1111410278-1377896170=:4191 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Vlad,=0AI am running in separate processes. But I think I found at least pa= rt of the problem.=0AIt is the HTable connection setup that is affecting it= =0AIf I setup all the connections in all the processes for all the tables t= hen and then call the endpoints it goes through ok.=0AAt least for the exam= ple below.=0ABut with my real code where one endpoint does some computation= of atleast 400-500msec. It is degrading exponentially.=0A(I will reduce th= e handler count and try ). Thanks and appreciate your response.=0AThese thi= ngs are blocking me from doing real scalability testing. We got single clie= nt down to 1.3 sec (target is under a sec) and now I want to move on to sca= lability.=0A=0ARegards,=0A- kiru=0A=0A=0A________________________________= =0A From: Vladimir Rodionov =0ATo: "user@hbase.apa= che.org" ; Kiru Pakkirisamy =0ASent: Friday, August 30, 2013 12:34 PM=0ASubject: RE: Coprocessor co= ncurrency =0A =0A=0ACan you check RS thread stack traces during your test? = If there are some concurrency/contention issues they will pop up in a stack= traces.=0AAlso, make sure that you run your client apps in a separate proc= esses.=0A=0ABest regards,=0AVladimir Rodionov=0APrincipal Platform Engineer= =0ACarrier IQ, www.carrieriq.com=0Ae-mail: vrodionov@carrieriq.com=0A=0A___= _____________________________________=0AFrom: Kiru Pakkirisamy [kirupakkiri= samy@yahoo.com]=0ASent: Friday, August 30, 2013 1:49 AM=0ATo: hbase mailing= list=0ASubject: Coprocessor concurrency=0A=0ASee below a null endpoint whi= ch takes 60-70msec on my 4 node ec2 cluster for a table with 45 regions.=0A= When I run 64 concurrent clients for this the latency jumps to 3000-3700 ms= ec.=0A=0A(zookeeper maxClientCnxs is set to 0 (unlimited) and hbase regions= erver handler count is 800). I hope I am not missing any configs for concur= rency)=0A=0A---------------------------------------------------------------= ---------------------------------------------------------------------------= ----------------=0A=0Apackage com.serendio.hbase.coprocessor;=0A=0Aimport j= ava.io.IOException;=0Aimport java.util.Map;=0A=0Aimport org.apache.hadoop.h= base.coprocessor.BaseEndpointCoprocessor;=0A=0A//Aggregation implementation= at a region.=0Apublic class SearchEndpoint extends BaseEndpointCoprocessor= implements=0ASearchProtocol {=0A@Override=0Apublic Map foo= (Map input, int topN)=0Athrows IOException {=0A// Implement= your logic here=0A// Map ret =3D new HashMap();=0A// ret.put("foo", 1.0);=0Areturn null;=0A}=0A}=0A=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=0A=0Apackage com.serendio.hbase.coprocess= or;=0A=0Aimport java.io.IOException;=0Aimport java.util.Map;=0A=0Aimport or= g.apache.hadoop.hbase.ipc.CoprocessorProtocol;=0A=0Apublic interface Search= Protocol extends CoprocessorProtocol {=0A=0Apublic Map foo(= Map input,=0Aint topN) throws IOException;=0A}=0A=0A=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=0Apackag= e com.serendio.hbase.coprocessor;=0A=0Aimport java.io.IOException;=0Aimport= java.util.HashMap;=0Aimport java.util.Map;=0A=0Aimport org.apache.hadoop.h= base.HBaseConfiguration;=0Aimport org.apache.hadoop.hbase.client.HTable;=0A= import org.apache.hadoop.hbase.client.HTableInterface;=0Aimport org.apache.= hadoop.hbase.client.coprocessor.Batch.Call;=0A=0Apublic class Search {=0Apu= blic static final String TEST_TABLE =3D "c4";=0Apublic static final String = TEST_FAMILY =3D "info";=0Apublic static final String TEST_QUALIFIER =3D "c1= ";=0A=0A/**=0A* ``=0A*=0A* @param args=0A*/=0Apublic static void main(Strin= g[] args) {=0A=0Aorg.apache.hadoop.conf.Configuration conf =3D HBaseConfigu= ration.create();=0AHTableInterface table =3D null;=0Atry {=0Atable =3D new = HTable(conf, TEST_TABLE);=0A=0A} catch (IOException e2) {=0A// TODO Auto-ge= nerated catch block=0Ae2.printStackTrace();=0A}=0A=0AMap> results =3D null;=0Afinal Map input =3D new Hash= Map();=0Ainput.put("test", 1.0);=0Afinal int topN =3D 10;= =0Along start =3D System.currentTimeMillis();=0Atry {=0Aresults =3D table.c= oprocessorExec(SearchProtocol.class, null, null,=0Anew Call>() {=0A=0A@Override=0Apublic Map call(= SearchProtocol instance)=0Athrows IOException {=0A// TODO Auto-generated me= thod stub=0Areturn instance.foo(input, topN);=0A}=0A});=0A} catch (IOExcept= ion e1) {=0A// TODO Auto-generated catch block=0Ae1.printStackTrace();=0A} = catch (Throwable e1) {=0A// TODO Auto-generated catch block=0Ae1.printStack= Trace();=0A}=0Along end =3D System.currentTimeMillis();=0ASystem.out.printl= n("Total search elapsed time: "=0A+ Long.toString(end - start));=0A}=0A=0A}= =0A=0A=0A=0ARegards,=0A- kiru=0A=0AConfidentiality Notice:=A0 The informati= on contained in this message, including any attachments hereto, may be conf= idential and is intended to be read only by the individual or entity to who= m this message is addressed. If the reader of this message is not the inten= ded recipient or an agent or designee of the intended recipient, please not= e that any review, use, disclosure or distribution of this message or its a= ttachments, in any form, is strictly prohibited.=A0 If you have received th= is message in error, please immediately notify the sender and/or Notificati= ons@carrieriq.com and delete or destroy any copy of this message and its at= tachments. --2017413661-1111410278-1377896170=:4191--