Return-Path: X-Original-To: apmail-couchdb-user-archive@www.apache.org Delivered-To: apmail-couchdb-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 939D1CCEF for ; Sat, 19 May 2012 07:37:15 +0000 (UTC) Received: (qmail 15882 invoked by uid 500); 19 May 2012 07:37:14 -0000 Delivered-To: apmail-couchdb-user-archive@couchdb.apache.org Received: (qmail 15770 invoked by uid 500); 19 May 2012 07:37:14 -0000 Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: user@couchdb.apache.org Delivered-To: mailing list user@couchdb.apache.org Received: (qmail 15753 invoked by uid 99); 19 May 2012 07:37:13 -0000 Received: from nike.apache.org (HELO nike.apache.org) (192.87.106.230) by apache.org (qpsmtpd/0.29) with ESMTP; Sat, 19 May 2012 07:37:13 +0000 X-ASF-Spam-Status: No, hits=3.6 required=5.0 tests=FROM_LOCAL_NOVOWEL,HK_RANDOM_ENVFROM,HK_RANDOM_FROM,HTML_MESSAGE,RCVD_IN_DNSWL_LOW,SPF_PASS X-Spam-Check-By: apache.org Received-SPF: pass (nike.apache.org: domain of cgsmcmlxxv@gmail.com designates 209.85.216.180 as permitted sender) Received: from [209.85.216.180] (HELO mail-qc0-f180.google.com) (209.85.216.180) by apache.org (qpsmtpd/0.29) with ESMTP; Sat, 19 May 2012 07:37:07 +0000 Received: by qcmv28 with SMTP id v28so2978770qcm.11 for ; Sat, 19 May 2012 00:36:46 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=yGynGY9ERpKbOzCqRSg+QADWEYLiCkhZLt95/6qLc1c=; b=X5xyXbQWhRFmgv6sxYfttx/M8iVT8bR34lV241QFMZCZxhAVQr8JKrlqbYwnlpOICn OTiW2dt8MT6g/sxHzt3PLyFtgoTKo3+tr410sK8Xt2jjb2/5QfLqwfkf7hE6NUBk7rA+ uandOxJ5HycdV1kyXZNXCnM/3352RUcmGWeu6SDEbiDQYsSA4d4YCZM886aFZP/nsX3d xKLX5ymtdGUud4Nkuf/REoulxBIhHN71W8LKsHna4RaXPKf8Ojw83fBfOJWQkdEJG/G0 R4aDP8IVddAV9ygWBjmR7zIOCk95usCAmihxF1iXgId3F54QGuarhSxf5QyP+ESMrKZQ Axaw== MIME-Version: 1.0 Received: by 10.224.196.200 with SMTP id eh8mr27340786qab.88.1337413006137; Sat, 19 May 2012 00:36:46 -0700 (PDT) Received: by 10.229.49.142 with HTTP; Sat, 19 May 2012 00:36:46 -0700 (PDT) In-Reply-To: <27FB150D-23A4-4993-B86F-5EDC0B820A36@apache.org> References: <4FB69C07.7070705@cedarsoft.com> <4FB6B7BC.9080403@cedarsoft.com> <27FB150D-23A4-4993-B86F-5EDC0B820A36@apache.org> Date: Sat, 19 May 2012 09:36:46 +0200 Message-ID: Subject: Re: Performance decrease from 1.0.1 to 1.2.0 From: CGS To: user@couchdb.apache.org Content-Type: multipart/alternative; boundary=20cf300fb33f26187a04c05ebb75 --20cf300fb33f26187a04c05ebb75 Content-Type: text/plain; charset=ISO-8859-1 As Adam mentioned, 1.2.0 is coming with the freedom to choose the options for the opened socket (the default options may not provide the best choice for your needs). To optimize the socket options for your needs, the possible options are given here: http://www.erlang.org/doc/man/inet.html#setopts-2 I know that this feature may confuse some users, but once you start to feel them, one can see the beauty of this feature. I do like this new feature and I am just a regular user. CGS On Fri, May 18, 2012 at 11:07 PM, Adam Kocoloski wrote: > It's a throughput versus latency thing. Disabling it is a pretty common > thing to do in modern networks. > > Adam > > On May 18, 2012, at 4:57 PM, Johannes Schneider wrote: > > > -----BEGIN PGP SIGNED MESSAGE----- > > Hash: SHA1 > > > > Thanks for the hint. Will try it out. > > But I am sure, there is a good reason for enabling that algorithm? > > > > > > On 18.05.2012 22:27, Adam Kocoloski wrote: > >> 40ms indicates Nagle's algorithm is kicking in and trying to reduce > >> the number of packets sent over the wire. Try setting the > >> following in the .ini configuration: > >> > >> socket_options = [{nodelay, true}] > >> > >> I believe that was the default in 1.0 and has changed since then. > >> Cheers, Adam > >> > >> On May 18, 2012, at 2:59 PM, Johannes Schneider wrote: > >> > >> Hi guys, > >> > >> > >> I am using Ubuntu 12.04 (x86_64) with the default CouchDB > >> installation. And I have created some very small tests to check the > >> performance. > >> > >> It is just one simple "get". > >> > >> > >> With 1.0.1 it used to take about 1 or 2 milli seconds. Since the > >> upgrade the same GET takes about 40 millis. I am quite sure that > >> there is some type of misconfiguration... Anybody has a hint for > >> me? Are there any special configuration options that changed from > >> 1.0.1 to 1.2.0 (nothing found in release notes)? Or did Canonical > >> some magic? > >> > >> > >> > >> Debug output in couch.log: [Fri, 18 May 2012 18:55:35 GMT] [debug] > >> [<0.400.0>] 'GET' > >> > /couch_unit_test_1337367299363_55042/_design/all/_view/byTypeAndId?startkey=%5B%22lens%22%5D&include_docs=true&endkey=%5B%22lens%22,%7B%7D%5D > >> > >> > > {1, > >> > >> > >> 1} from "127.0.0.1" Headers: [{'Accept',"text/html, image/gif, > >> image/jpeg, *; q=.2, */*; q=.2"}, {'Connection',"keep-alive"}, > >> {'Host',"localhost:5984"}, {'User-Agent',"Java/1.6.0_26"}] [Fri, 18 > >> May 2012 18:55:35 GMT] [debug] [<0.400.0>] OAuth Params: > >> [{"startkey","[\"lens\"]"}, {"include_docs","true"}, > >> {"endkey","[\"lens\",{}]"}] [Fri, 18 May 2012 18:55:35 GMT] [debug] > >> [<0.400.0>] request_group {Pid, Seq} {<0.402.0>,5} [Fri, 18 May > >> 2012 18:55:35 GMT] [info] [<0.400.0>] 127.0.0.1 - - GET > >> > /couch_unit_test_1337367299363_55042/_design/all/_view/byTypeAndId?startkey=%5B%22lens%22%5D&include_docs=true&endkey=%5B%22lens%22,%7B%7D%5D > >> > >> > > 200 > >> > >> > >> The map function. ------------------------------ function( doc ) { > >> var type = doc['@type']; var id = doc['_id']; > >> > >> if ( type == null || id == null ) { return } > >> > >> emit( [type, id] ); } ------------------------------ > >> > >> > >> No other documents are available (returns "total_rows":0 > >> > >> > >> > >> Thanks, > >> > >> Johannes > >> > >> > > > > - -- > > Johannes Schneider - blog.cedarsoft.com > > -----BEGIN PGP SIGNATURE----- > > Version: GnuPG v1.4.11 (GNU/Linux) > > > > iQEcBAEBAgAGBQJPtre8AAoJEAytD9R7Qv6dcKEH/RhHUVBbyJsUSUK5vx4eD5sR > > GBIQrvvAHqLiQ9eQ8jH5jucvF3uhX8dBHEID19IbGDmpGbzk+LQtNtkKsaC5844J > > pOzJDC1ObvYwtQFaPeUX3mNkV7nkKwrIoB5x1zB1/4utQmlPMkxCrv4XlJ79DcPx > > 1m59DItFzCTSLUxzG1r1/g3u93xAsxPpBMp0IXFseuXOmuiO5v64u7hXriN3eSWT > > 6PVhz88AzJTZlYH+10uMuMPbaxcosgNOtbVJDPGXXysziuHEe4gpJB3fQwbs1uow > > cmVdhB2Ln8I/MJzuGaKwtBncLPnafPSdYwlhsOBEwSVbc5tEDw2wkLKwHBtQsVk= > > =b6eZ > > -----END PGP SIGNATURE----- > > --20cf300fb33f26187a04c05ebb75--