Age | Commit message (Collapse) | Author | |
---|---|---|---|
2010-10-20 | let rebar manage the module lists | Adam Kocoloski | |
2010-10-20 | repair should run with R=N | Adam Kocoloski | |
2010-10-19 | use new read repair and ancestry checks in fabric:open_doc/3 | Adam Kocoloski | |
2010-10-19 | remove_ancestors/2 will be generally useful | Adam Kocoloski | |
2010-10-19 | fix quorum check and be sure repair runs when revs are missing w/ open_revs=all | Adam Kocoloski | |
2010-10-19 | fix DB compaction failing to switch over. Reported by rnewson | Adam Kocoloski | |
2010-10-19 | suppress unused variable warning | Adam Kocoloski | |
2010-10-19 | improved read quorum and repair for open_revs | Adam Kocoloski | |
There are important distinctions between requests with open_revs=all, requests with an explicit revision list and latest=true, and a simple revision list. When open_revs=all, we take the union of all revisions returned by the workers, then reduce it by filtering out ancestors on each edit branch. An ancestor counts towards the descendant's read quorum. If an explicit revision list is given, a worker can respond with {{not_found, missing}, Rev}, an {ok, #doc{}} corresponding to the rev in question, or an {ok, #doc{}} with a newer revision if latest=true is supplied. not_found is considered to be an ancestor of any {ok,_} revision for quorum purposes. Down workers do not contribute to the read quorum. If workers disagree at all on the response, read repair will run asynchronously. The repair tries to save all the latest #doc{} records (i.e. the ones w/o descendants) which were received during the voting. BugzID 11047 | |||
2010-10-01 | minor tweak from Adam to fix replication, change method spec to correct ↵ | Robert Dionne | |
method for view_cleanup | |||
2010-09-13 | accept logins with 'name' instead of 'username', thanks artemb | Adam Kocoloski | |
2010-09-06 | remove fail_on_warning, closes #2 | Adam Kocoloski | |
Warnings appear to be unavoidable if we want to support R13B03 and R14A | |||
2010-09-06 | support for filtered _changes on multi-node clusters, closes #5 | Adam Kocoloski | |
2010-09-03 | suppress http:request/4 deprecation warning. Closes #2 | Adam Kocoloski | |
2010-09-03 | s/Cloudant/BigCouch | Adam Kocoloski | |
2010-09-03 | support for _view_cleanup via HTTP | Adam Kocoloski | |
2010-09-01 | pass W from Req to fabric properly | Brad Anderson | |
2010-09-01 | whitespace | Brad Anderson | |
2010-08-29 | fix heartbeat for _changes feeds | Adam Kocoloski | |
2010-08-27 | use couchspawnkillable from privdir instead of overlay | Adam Kocoloski | |
2010-08-27 | add README.md back in for apps | Brad Anderson | |
2010-08-27 | my name is BigCouch | Adam Kocoloski | |
2010-08-27 | split some rexi utilities out from fabric | Brad Anderson | |
2010-08-27 | use dbcore not core | joewilliams | |
2010-08-27 | fix a bad cherry-pick | Adam Kocoloski | |
2010-08-27 | display full cluster update_seq in get_db_info() | Adam Kocoloski | |
2010-08-27 | fix "Mike's bug" with attachment uploading. BugzID 10742 | Adam Kocoloski | |
2010-08-27 | avoid deadlocks when changing query_servers configs, BugzID 10679 | Adam Kocoloski | |
2010-08-27 | tolerate view groups initializing during DB deletion | Adam Kocoloski | |
2010-08-27 | implement cleanup_index_files() | Adam Kocoloski | |
2010-08-27 | slightly better rexi_EXIT handling, needs more work | Adam Kocoloski | |
2010-08-27 | server and client should both update LRU table, BugzID 10755 | Adam Kocoloski | |
2010-08-27 | return well-formed response for reduce view with 0 rows, BugzID 10725 | Adam Kocoloski | |
2010-08-27 | resume reducers after the row is sent, BugzID 10683 | Adam Kocoloski | |
2010-08-27 | initialize #db{} so that validation funs are loaded on first write | Adam Kocoloski | |
BugzID 10675 | |||
2010-08-27 | Apache 2 license, Cloudant copyright when appropriate | Adam Kocoloski | |
2010-08-27 | remove some logging | Adam Kocoloski | |
2010-08-26 | log to var/log/dbcore.log by default | Adam Kocoloski | |
2010-08-23 | consolidate all couchjs stuff in one subdir | Adam Kocoloski | |
2010-08-20 | a bit more cleanup of root dir | Brad Anderson | |
2010-08-20 | build couchjs with scons | Brad Anderson | |
2010-08-19 | make dist back in working order, and dbcore starts successfully | Brad Anderson | |
2010-08-19 | Cloudant changes to etap tests | Brad Anderson | |
2010-08-19 | etap couch tests now executing again, work to do on C libs to get them to pass | Brad Anderson | |
2010-08-18 | disable ibrowse test for now | Brad Anderson | |
2010-08-18 | lay down cloudant build system. | Brad Anderson | |
2010-08-18 | clear out root folder a bit, moving couch bits into apps/couch or rel/ | Brad Anderson | |
2010-08-18 | new version() call allows for hidden patch releases | Adam Kocoloski | |
2010-08-18 | design_doc_to_view_group only needs the #doc{} | Adam Kocoloski | |
2010-08-18 | Grab bag of Cloudant patches to couch OTP application | Adam Kocoloski | |
- Removal of couch_db and couch_ref_counter processes. Active DBs are accessible through a protected ets table owned by couch_server. - #full_doc_info{} in by_id and by_seq trees for faster compaction at the expense of more disk usage afterwards. Proposed as COUCHDB-738 but not accepted upstream. - Replication via distributed Erlang. - Better hot upgrade support (uses exported functions much more often). - Configurable btree chunk sizes allow for larger (but still bounded) reductions. - Shorter names for btree fields in #db{} and #db_header{}. - couch_view_group does not keep a reference to the #db{}. - Terms are stored compressed (again). | |||
2010-08-18 | couch_util:to_binary now catches all errors | Adam Kocoloski | |