[Couchdb Wiki] Update of Installing_on_Ubuntu by paulbarrett
Dear Wiki user, You have subscribed to a wiki page or wiki category on Couchdb Wiki for change notification. The Installing_on_Ubuntu page has been changed by paulbarrett. The comment on this change is: Changes to get 0.11 and 1.0 running because things are missing in make. http://wiki.apache.org/couchdb/Installing_on_Ubuntu?action=diffrev1=36rev2=37 -- # wget url tar xvzf apache-couchdb-x.xx.x.tar.gz cd apache-couchdb-x.xx.x + + # see xulrunner directions below and do them this fixes problem where basic test suite fails to run. # Note: To install couchdb in the default location use --prefix= in the configure statement ./configure --prefix= --with-js-lib=/usr/lib/xulrunner-devel-1.9.2.6/lib --with-js-include=/usr/lib/xulrunner-devel-1.9.2.6/include + # Now you can compile and install couchdb make make install # Add couchdb user account useradd -d /var/lib/couchdb couchdb chown -R couchdb: /var/lib/couchdb /var/log/couchdb + + # next two steps fix problems where adding admin hangs or setting admins in local.ini hangs the start. Also fixes problems with reader_acl test. + chown -R root:couchdb /etc/couchdb + chmod -R 664 /etc/couchdb + # start couchdb /etc/init.d/couchdb start # Start couchdb on system start
[Couchdb Wiki] Update of Installing_on_Ubuntu by afters
Dear Wiki user, You have subscribed to a wiki page or wiki category on Couchdb Wiki for change notification. The Installing_on_Ubuntu page has been changed by afters. The comment on this change is: couch_v 0.11.0 - 1.0.0. http://wiki.apache.org/couchdb/Installing_on_Ubuntu?action=diffrev1=38rev2=39 -- # {couchdb:Welcome,version:0.11.0} }}} === Example 2 (with minimal dependencies) === - Tested with CouchDB versions: 0.11, 1.0.0 + Tested with CouchDB versions: 0.11.0, 1.0.0 This example is especially useful when using a minimal Ubuntu, rather than a full desktop one, as it won't install as many packages as the previous example. @@ -92, +92 @@ '''Note''': Without the 'erlang' package, couchdb could still be installed, but would be missing important erlang libraries and won't run properly. Installation: {{{ - couch_v='0.11.0' + couch_v='1.0.0' prefix='/usr/local' # install:
svn commit: r964956 - in /couchdb/trunk/src/couchdb: couch_db.erl couch_rep_att.erl
Author: fdmanana Date: Fri Jul 16 21:32:57 2010 New Revision: 964956 URL: http://svn.apache.org/viewvc?rev=964956view=rev Log: Fix for a pull replication, targeted to a 1.0 CouchDB server, where the source DB is in a remote CouchDB 0.11.0 server and the target DB is local (1.0 CouchDB DB). Closes ticket COUCHDB-827. Modified: couchdb/trunk/src/couchdb/couch_db.erl couchdb/trunk/src/couchdb/couch_rep_att.erl Modified: couchdb/trunk/src/couchdb/couch_db.erl URL: http://svn.apache.org/viewvc/couchdb/trunk/src/couchdb/couch_db.erl?rev=964956r1=964955r2=964956view=diff == --- couchdb/trunk/src/couchdb/couch_db.erl (original) +++ couchdb/trunk/src/couchdb/couch_db.erl Fri Jul 16 21:32:57 2010 @@ -901,10 +901,20 @@ with_stream(Fd, #att{md5=InMd5,type=Type write_streamed_attachment(_Stream, _F, 0) - ok; +% LenLeft might be different from the total size of what function F returns. +% This happens when doing a pull replication of compressed attachments from a +% 0.11.0 server, where LenLeft will match the uncompressed size but we end up +% receiving the attachment compressed (therefore a size different from LenLeft). +% This is because 0.11.0 doesn't understand the query parameter +% ?att_encoding_info=true when we do a doc request (GET /somedb/somedoc). write_streamed_attachment(Stream, F, LenLeft) - -Bin = F(), -ok = couch_stream:write(Stream, Bin), -write_streamed_attachment(Stream, F, LenLeft - size(Bin)). +case F() of +Bin when is_binary(Bin) - +ok = couch_stream:write(Stream, Bin), +write_streamed_attachment(Stream, F, LenLeft - size(Bin)); +eof - +ok +end. enum_docs_since_reduce_to_count(Reds) - couch_btree:final_reduce( Modified: couchdb/trunk/src/couchdb/couch_rep_att.erl URL: http://svn.apache.org/viewvc/couchdb/trunk/src/couchdb/couch_rep_att.erl?rev=964956r1=964955r2=964956view=diff == --- couchdb/trunk/src/couchdb/couch_rep_att.erl (original) +++ couchdb/trunk/src/couchdb/couch_rep_att.erl Fri Jul 16 21:32:57 2010 @@ -81,8 +81,12 @@ receive_data(Ref, ReqId, ContentEncoding % ?LOG_DEBUG(got ~p bytes for ~p, [size(Data), ReqId]), Data; {ibrowse_async_response_end, ReqId} - -?LOG_ERROR(streaming att. ended but more data requested ~p, [ReqId]), -throw({attachment_request_failed, premature_end}) +% This means ibrowse received all the data it was supposed to. +% In case of not receiving the whole data, due to a network link +% failure for example, we would have received an error message. +% In other words, this message doesn't represent an error - look into +% ibrowse_http_client.erl. +eof after 31000 - throw({attachment_request_failed, timeout}) end.
svn commit: r964971 - in /couchdb/branches/1.0.x/src/couchdb: couch_db.erl couch_rep_att.erl
Author: fdmanana Date: Fri Jul 16 22:56:55 2010 New Revision: 964971 URL: http://svn.apache.org/viewvc?rev=964971view=rev Log: Revert revision 964957 (COUCHDB-827) - solution had secondary effects. Modified: couchdb/branches/1.0.x/src/couchdb/couch_db.erl couchdb/branches/1.0.x/src/couchdb/couch_rep_att.erl Modified: couchdb/branches/1.0.x/src/couchdb/couch_db.erl URL: http://svn.apache.org/viewvc/couchdb/branches/1.0.x/src/couchdb/couch_db.erl?rev=964971r1=964970r2=964971view=diff == --- couchdb/branches/1.0.x/src/couchdb/couch_db.erl (original) +++ couchdb/branches/1.0.x/src/couchdb/couch_db.erl Fri Jul 16 22:56:55 2010 @@ -901,20 +901,10 @@ with_stream(Fd, #att{md5=InMd5,type=Type write_streamed_attachment(_Stream, _F, 0) - ok; -% LenLeft might be different from the total size of what function F returns. -% This happens when doing a pull replication of compressed attachments from a -% 0.11.0 server, where LenLeft will match the uncompressed size but we end up -% receiving the attachment compressed (therefore a size different from LenLeft). -% This is because 0.11.0 doesn't understand the query parameter -% ?att_encoding_info=true when we do a doc request (GET /somedb/somedoc). write_streamed_attachment(Stream, F, LenLeft) - -case F() of -Bin when is_binary(Bin) - -ok = couch_stream:write(Stream, Bin), -write_streamed_attachment(Stream, F, LenLeft - size(Bin)); -eof - -ok -end. +Bin = F(), +ok = couch_stream:write(Stream, Bin), +write_streamed_attachment(Stream, F, LenLeft - size(Bin)). enum_docs_since_reduce_to_count(Reds) - couch_btree:final_reduce( Modified: couchdb/branches/1.0.x/src/couchdb/couch_rep_att.erl URL: http://svn.apache.org/viewvc/couchdb/branches/1.0.x/src/couchdb/couch_rep_att.erl?rev=964971r1=964970r2=964971view=diff == --- couchdb/branches/1.0.x/src/couchdb/couch_rep_att.erl (original) +++ couchdb/branches/1.0.x/src/couchdb/couch_rep_att.erl Fri Jul 16 22:56:55 2010 @@ -81,12 +81,8 @@ receive_data(Ref, ReqId, ContentEncoding % ?LOG_DEBUG(got ~p bytes for ~p, [size(Data), ReqId]), Data; {ibrowse_async_response_end, ReqId} - -% This means ibrowse received all the data it was supposed to. -% In case of not receiving the whole data, due to a network link -% failure for example, we would have received an error message. -% In other words, this message doesn't represent an error - look into -% ibrowse_http_client.erl. -eof +?LOG_ERROR(streaming att. ended but more data requested ~p, [ReqId]), +throw({attachment_request_failed, premature_end}) after 31000 - throw({attachment_request_failed, timeout}) end.
svn commit: r964987 - /couchdb/trunk/share/www/couch_tests.html
Author: jchris Date: Sat Jul 17 00:11:00 2010 New Revision: 964987 URL: http://svn.apache.org/viewvc?rev=964987view=rev Log: note about firefox support in test suite Modified: couchdb/trunk/share/www/couch_tests.html Modified: couchdb/trunk/share/www/couch_tests.html URL: http://svn.apache.org/viewvc/couchdb/trunk/share/www/couch_tests.html?rev=964987r1=964986r2=964987view=diff == --- couchdb/trunk/share/www/couch_tests.html [utf-8] (original) +++ couchdb/trunk/share/www/couch_tests.html [utf-8] Sat Jul 17 00:11:00 2010 @@ -68,7 +68,10 @@ specific language governing permissions strongNote:/strong Each of the tests will block the browser. If the connection to your CouchDB server is slow, running the tests will take some time, and you'll not be able to do much with your browser while -a test is being executed. +a test is being executed. strongAlso:/strong The test suite is designed +to work with Firefox (with Firebug disabled). Patches are welcome for +convenience compatibility with other browsers, but official support is +for Firefox (latest stable version) only. /p table class=listing id=tests cellspacing=0