Age | Commit message (Collapse) | Author |
|
|
|
Our headers start with a <<1>> and then four bytes indicating the length
of the header and its checksum. When the header is larger than 4090
bytes it will be split across multiple blocks in the file and will need
to be reassembled on read. The reassembly consists of stripping out
<<0>> from the beginning of each subsequent block in the
remove_block_prefixes/2 function. The bug here is that we tell
remove_block_prefixes that we're starting 1 byte into the current block
instead of 5, so it ends up removing one good byte from the header and
injecting one or more random <<0>>s.
Headers larger than 4k are very rare and generally require a view group
with a huge number of indexes or indexes with fairly large reductions,
which explains why this bug has gone undetected until now.
Closes COUCHDB-1319.
|
|
Looks like we had a couple commits get crossed over during the merges.
This removes the original version of the patch that wrapped JS functions
with parens.
|
|
|
|
|
|
This reverts commit e311d45fba617e7917d76109b4e21dbb388f268d.
Turns out the 'libm' error is a very obscure way of reporting a lack
of a c++ compiler.
|
|
BugzID:12741
|
|
|
|
|
|
Add a remsh script for each node
BugzId: 12730
|
|
Cookies now default to using ~/.erlang.cookie which is automatically
generated if it does not exist when erlang starts with a name.
There is also no longer a need to specify a name to remsh as it'll
generate a name based on the process id and set the short or long
version appropriately.
|
|
Fix badarg error in couch_server:try_close_lru/1
BugzID: 12363
|
|
The race condition in couch_server's ets table usage rears its ugly head
by leaving an entry in couch_lru. This patch just addresses the issue by
allowing the client pid to use the db and ignores the fact that for the
duration its over the max_dbs_open setting.
|
|
|
|
|
|
The remsh pattern is quite useful for connecting to running nodes for
debugging production issues. This adds a small script that lets users
connect quickly to a running node.
|
|
Part of the backport of COUCHDB-1008 to chttpd. This just updates
the default.ini with some commented out options describing their
availability to users.
|
|
This function is new in Python 2.7 and we already have a function that
returns command ouput. Just reuse runcmd to get the version.
|
|
|
|
This fixes "Unable to generate spec" errors when trying to build a
release on Erlang/OTP R13.
|
|
This reverts commit 61f10e7590bd352b367ad426a56587cbef3700b6.
|
|
This reverts commit 5b558c81ed9709fb286a6821e9ae6d6478012c2c.
|
|
|
|
This commit reverts 1.8.5 compatibility including the corresponding couchjs
paren hack as this leads to significant breakage in existing functions.
|
|
|
|
resolved in 1.1.1
|
|
|
|
|
|
|
|
|
|
This change makes argument parsing in couchjs more robust:
* Removes the need for couchjs script
* Adds long options
* Makes all the options to couchjs available (fix COUCHDB-893)
Conflicts:
bin/Makefile.am
bin/couchjs.tpl.in
configure.ac
|
|
Test by Christian Carter
This is a backport of r1177890
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1177892 13f79535-47bb-0310-9956-ffa450edef68
|
|
Due to a bug, older releases allowed the creation of
documents with an empty ID, which are impossible to
GET therefore making the replicator crash.
This change simply skips such documents and logs
an error message to inform users.
This is a backport of revision 1177548 from trunk.
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1177550 13f79535-47bb-0310-9956-ffa450edef68
|
|
This module is deprecated in OTP R15, which is going to
be released by the end of this year.
The etap tests now use ibrowse instead.
This is a backport of revision 1177459 from trunk.
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1177463 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
I was slaving away on this merge for two days.
|
|
|
|
Conflicts:
apps/couch/include/couch_db.hrl
apps/couch/src/couch_db.erl
apps/couch/src/couch_os_process.erl
apps/couch/src/couch_query_servers.erl
apps/couch/src/couch_rep.erl
apps/couch/src/couch_replication_manager.erl
apps/couch/src/couch_view_compactor.erl
apps/couch/src/couch_view_group.erl
apps/couch/src/couch_view_updater.erl
configure.ac
couchjs/c_src/http.c
couchjs/c_src/main.c
couchjs/c_src/utf8.c
etc/windows/couchdb.iss.tpl
src/couchdb/priv/Makefile.am
src/couchdb/priv/couch_js/main.c
test/etap/160-vhosts.t
test/etap/200-view-group-no-db-leaks.t
test/etap/Makefile.am
BugzID: 12645
|
|
As a follow up to COUCHDB-1265 I was missing the fact that after the
insertion of a new update_seq into an internal node it is quite possible
that a compaction runs before the doc is updated again. This is
important because compaction removes information of the largest update
seq from the tree itself.
The fix is simple to include the update_seq from the #full_doc_info{}
record when calculating #doc_info.high_seq. The way to think of this
is that it's the maximum value from all known values for the update
sequence which can be defined as all values known in the tree or in the
full_doc_info record.
Backport of r1176701 from trunk
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1176704 13f79535-47bb-0310-9956-ffa450edef68
|
|
We were making a typical JavaScript mistake testing a module for
truthiness instead of checking that it was undefined.
Fixes COUCHDB-1223
Backport of r1176677 from trunk
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1176679 13f79535-47bb-0310-9956-ffa450edef68
|
|
Found this error using the Debian package for SM 1.8.5 and have since
had reports of users seeing it as well. The basic error is that some
versions of SpiderMonkey appear to dislike this call to eval:
eval("function(){}");
The fix is simply to wrap the function source in parenthesis so that
SM is convinced that it knows how to evaluate a function.
Backport of r1176666 from trunk.
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1176667 13f79535-47bb-0310-9956-ffa450edef68
|
|
Thanks to Dave Cottlehuber for the patch.
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1176651 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
Writing the header first ensures that the index will not be reset if
the server crashes in between the handoff and the next delayed_commit.
See COUCHDB-994
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1174979 13f79535-47bb-0310-9956-ffa450edef68
|
|
This is tested against the 1.7.0, 1.8.0rc1, and 1.8.5 tarballs from
Mozilla's FTP directory. It's mostly the same code from trunk minus a
few tweaks to get it past a couple type errors using c++ instead of cc.
|
|
fix port conversion, don't forget to use substr function. spotted by
@rnewson on irc, again. thanks. patch by benoitc.
git-svn-id: https://svn.apache.org/repos/asf/couchdb/branches/1.1.x@1174358 13f79535-47bb-0310-9956-ffa450edef68
|