Return-Path: Delivered-To: apmail-couchdb-user-archive@www.apache.org Received: (qmail 24584 invoked from network); 5 Aug 2010 18:37:11 -0000 Received: from unknown (HELO mail.apache.org) (140.211.11.3) by 140.211.11.9 with SMTP; 5 Aug 2010 18:37:11 -0000 Received: (qmail 84168 invoked by uid 500); 5 Aug 2010 18:37:10 -0000 Delivered-To: apmail-couchdb-user-archive@couchdb.apache.org Received: (qmail 84117 invoked by uid 500); 5 Aug 2010 18:37:09 -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 84101 invoked by uid 99); 5 Aug 2010 18:37:09 -0000 Received: from nike.apache.org (HELO nike.apache.org) (192.87.106.230) by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 05 Aug 2010 18:37:09 +0000 X-ASF-Spam-Status: No, hits=-0.0 required=10.0 tests=RCVD_IN_DNSWL_LOW,SPF_NEUTRAL X-Spam-Check-By: apache.org Received-SPF: neutral (nike.apache.org: local policy) Received: from [209.68.5.16] (HELO relay02.pair.com) (209.68.5.16) by apache.org (qpsmtpd/0.29) with SMTP; Thu, 05 Aug 2010 18:37:02 +0000 Received: (qmail 8917 invoked from network); 5 Aug 2010 18:36:37 -0000 Received: from 74.1.186.35 (HELO ?10.0.1.205?) (74.1.186.35) by relay02.pair.com with SMTP; 5 Aug 2010 18:36:37 -0000 X-pair-Authenticated: 74.1.186.35 Content-Type: text/plain; charset=us-ascii Mime-Version: 1.0 (Apple Message framework v1081) Subject: Re: View Building Grinds System to Halt From: Damien Katz In-Reply-To: Date: Thu, 5 Aug 2010 11:36:36 -0700 Content-Transfer-Encoding: quoted-printable Message-Id: <9DDC8E1B-D5BD-4074-839B-34CD0D7566FC@apache.org> References: To: user@couchdb.apache.org X-Mailer: Apple Mail (2.1081) X-Virus-Checked: Checked by ClamAV on apache.org Can you post your design doc with the views in it? Keep in mind, CouchDB = builds all views in a design doc at the same time, so it might not be = the view you are accessing that's using up resources, but another = included in the same design doc. -Damien On Aug 5, 2010, at 11:29 AM, Gordon Leland Hempton wrote: > This also seems to happen whenever I cleanup views: beam.smp uses all = the > systems memory and then disappears, however the log message doesn't = show up. >=20 > Views work totally fine on databases that are small in size. >=20 > Anyone have any recommendations? >=20 > On Wed, Aug 4, 2010 at 4:07 PM, Gordon Leland Hempton = wrote: >=20 >> After looking at the logs, it looks pretty bad: >>=20 >> [Wed, 04 Aug 2010 22:57:28 GMT] [info] [<0.12883.0>] 127.0.0.1 - - = 'GET' >> = /trademark/_design/Trademark/_view/by_filingDate?limit=3D10&stale=3Dok&des= cending=3Dtrue&include_docs=3Dtrue >> 200 >>=20 >> [Wed, 04 Aug 2010 22:57:52 GMT] [error] [emulator] Error in process >> <0.12957.0> with exit value: = {badarg,[{erlang,binary_to_term,[<<1885695232 >> = bytes>>]},{couch_file,pread_term,2},{couch_db,make_doc,5},{couch_db,open_d= oc_int,3},{couch_view_updater,load_doc,5},{couch_view_updater,'-update/2-f= un-2-',8},{couch_btree,stream_kv_node2,8... >>=20 >> Anyone know where to start? >>=20 >> I am running Ubuntu 10.04 server with couchdb 1.0 installed from = source. I >> have made sure the xulrunner settings are correct. >>=20 >> Thanks, >>=20 >> On Wed, Aug 4, 2010 at 12:18 PM, Gordon Leland Hempton = >> wrote: >>=20 >>> I am working with a database of ~10 mil docs and 30GB size. I am = trying to >>> build a view consisting of a single field as the key and a null = document. >>>=20 >>> After creating the view and trying to access it, the couchdb = beam.smp >>> command eventually goes up to 12.3GB virtual memory and uses 3.7GB = of >>> physical memory. Memory is clearly the bottleneck as the CPU usage = is around >>> 5-15%. My system only has 4GB of memory so naturally all this paging = is >>> slowing everything down to a crawl. >>>=20 >>> After a few minutes this usage will go back down, but it will get to = the >>> same levels every time I access the view (which is still building = and has >>> been for 12 hours). >>>=20 >>> Is this normal? How can I improve this? >>>=20 >>> -- >>> Gordon L. Hempton >>> http://componentlab.com >>> http://hempton.com >>>=20 >>=20 >>=20 >>=20 >> -- >> Gordon L. Hempton >> http://componentlab.com >> http://hempton.com >>=20 >=20 >=20 >=20 > --=20 > Gordon L. Hempton > http://componentlab.com > http://hempton.com