Permalink
Commits on Nov 7, 2011
  1. @agentzh
  2. @agentzh
  3. @agentzh

    * bugfix: we should no longer return NGX_AGAIN when the re-polling

      returns IO WAIT in case of the "connection made" state.
    * added a test for reproducing a stale event handling bug at least in
      ngx_epoll_module.
    * introduced the TEST_NGINX_POSTGRESQL_HOST environment to specify a
      remote Pg server to test against.
    agentzh committed Nov 7, 2011
Commits on Nov 4, 2011
  1. @agentzh

    bugfix: now we log an error message when the postgres_pass target is …

    …not found at all and returns 500 in this case instead of returning empty response.
    agentzh committed Nov 4, 2011
  2. @agentzh

    fixed another failing test in t/rewrites.t for pg 9.x. but the rewrit…

    …es.t is still failing due to the "postgres_rewrite changes" feature assumes the rows_affected slot to mean "rows changed", but this is no longer true in pg 9.x. hopefully Piotr Sikora can address this issue for us :)
    agentzh committed Nov 4, 2011
Commits on Sep 25, 2011
  1. @agentzh

    updated the test suite for postgres 9.0.x: affected_rows take non-zer…

    …o values for select queries.
    agentzh committed Sep 25, 2011
  2. @agentzh

    fixed compatibility issues with nginx 1.1.4: ngx_chain_update_chains …

    …now requires a pool argument.
    agentzh committed Sep 25, 2011
Commits on Aug 4, 2011
  1. @agentzh

    confirmed that we work with nginx 1.0.5; also added the Installation …

    …section to READE, especially documented the LIBPQ_INC and LIBPQ_LIB environments.
    agentzh committed Aug 4, 2011
Commits on Jul 13, 2011
  1. @agentzh

    fixed an issue regarding defining global variables in C header files:…

    … we should have defined rds_rough_col_type_t as a type rather than a global variable. thanks @姜大炮.
    agentzh committed Jul 13, 2011
Commits on Jun 21, 2011
  1. @PiotrSikora

    Always return first (and only) value for "postgres_output value".

    This is second part of the "write proper SQL queries" campaign. Queries
    that return more than one value will result in "500 Internal Server Error"
    response.
    PiotrSikora committed Jun 21, 2011
  2. @PiotrSikora

    Replace "postgres_output row <row>" with "postgres_output text".

    This is more general approach, which forces writing proper SQL queries
    instead of filtering results on the nginx side and allows for sending
    output from multiple rows to end-users.
    
    Discussed with Silly Sad.
    PiotrSikora committed Jun 21, 2011
  3. @PiotrSikora
Commits on Jun 17, 2011
  1. @PiotrSikora

    Improve library discovery process by using PostgreSQL's pg_config.

    Based on patch from Silly Sad.
    PiotrSikora committed Jun 17, 2011
  2. @PiotrSikora

    Don't set last_buf=1 (duplicated last chunk issue).

    Because ngx_http_upstream_finalize_request sends NGX_HTTP_LAST,
    having last_buf=1 in our module meant that last two chains always
    had last_buf=1, which resulted in duplicated last chunk sent for
    HTTP/1.1 requests. This pretty much killed keep-alived requests.
    
    Reported by Silly Sad, diagnosed by Maxim Dounin.
    
    Same issue was independently diagnosed and fixed in ngx_drizzle
    by Yichun Zhang (agentzh) few days ago.
    PiotrSikora committed Jun 17, 2011
Commits on May 11, 2011
  1. @agentzh
  2. @agentzh

    updated .gitignore a bit.

    agentzh committed May 11, 2011
Commits on Feb 18, 2011
  1. @agentzh
  2. @agentzh

    removed the hand-written << "in upstream \"t.shV\"", &pgdt->name >> p…

    …art from most of the ngx_log_error calls in ngx_postgres_keepalive.c because they are duplicate with the << upstream: "postgres://ip.add.re.ss:port" >> part automatically generated by the ngx_http_log_error_handler function in the nginx core.
    agentzh committed Feb 18, 2011
  3. @agentzh

    fixed a bug in the connection pool: we should copy over the "name" an…

    …d "sockaddr" fields instead of passing pointers around.
    agentzh committed Feb 18, 2011
Commits on Feb 17, 2011
  1. @agentzh

    fixed a minor bug in the connection pool: we should resume the "name"…

    …, "sockaddr", and "socklen" fields for the connection from the pool such that we can get more detailed error log messages with the "upstream: postgres://ip.add.re.ss:port" bit.
    agentzh committed Feb 17, 2011
Commits on Jan 13, 2011
  1. @PiotrSikora

    Use PQescapeLiteral() to detect PostgreSQL 9.0.

    Calling PQconnectStartParams(NULL, *, *) results in SIGSEGV.
    PiotrSikora committed Jan 13, 2011
Commits on Jan 4, 2011
  1. @PiotrSikora
Commits on Dec 23, 2010
  1. @PiotrSikora
  2. @PiotrSikora
  3. @PiotrSikora
  4. @PiotrSikora
  5. @PiotrSikora
  6. @PiotrSikora

    Revert fixed jump on uninitialized value.

    The problem is that this value shouldn't be used in the first place
    (see next commit).
    
    This reverts commit 1ed9f15.
    PiotrSikora committed Dec 23, 2010
  7. @PiotrSikora
  8. @PiotrSikora
  9. @PiotrSikora

    Revert "postgres_binary_mode".

    This is great feature, but the way it hooks into configuration must
    be rewritten (as discussed with Yichun on the IRC).
    
    This reverts commit 4187c57.
    PiotrSikora committed Dec 23, 2010
Commits on Dec 22, 2010
  1. @PiotrSikora
  2. @PiotrSikora

    Revert fix backported from ngx_drizzle.

    I've just tested each case and there is no such bug in ngx_postgres,
    each error.log entry contains correct upstream peer name.
    
    I'll re-add "pc->name = &peer->name;" in case nginx or 3rd-party
    modules will ever use it, but "pc->name" isn't used by ngx_postgres.
    
    This reverts commit 8126a4f.
    PiotrSikora committed Dec 22, 2010
  3. @PiotrSikora
  4. @PiotrSikora