Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
branch: master
Commits on Mar 3, 2015
  1. Andrew Gwozdziewycz

    Update documentation to express the true nature

    apg authored
    application/logplex-1 really is inspired by RFC6587, not RFC5424.
Commits on Feb 5, 2015
  1. Alex Arnell
  2. Alex Arnell

    adding upgrade script

    voidlock authored
  3. Alex Arnell

    add comment describing issue

    voidlock authored
  4. Alex Arnell
Commits on Feb 4, 2015
  1. Alex Arnell
  2. Alex Arnell

    adding live_upgrade script

    voidlock authored
Commits on Jan 30, 2015
  1. Alex Arnell

    remove commented out code

    voidlock authored
  2. Alex Arnell
  3. Alex Arnell

    fixup compiler issues

    voidlock authored
  4. Alex Arnell
  5. Alex Arnell

    let API append this parameter

    voidlock authored
  6. Alex Arnell
  7. Phil Hagelberg Alex Arnell

    Return absolute URLs for tail sessions.

    technomancy authored voidlock committed
Commits on Jan 26, 2015
  1. Alex Arnell

    adding live_upgrade script

    voidlock authored
  2. Alex Arnell

    revert hibernate changes

    voidlock authored
    Monitoring of the logplex cluster revealed that post this change set the
    binary memory usage climbed. IMO this doesn't make sense because the
    drain buffer process will still retain references to any binaries
    cleaned up by forcing a hibernate here. But, graphs do not lie. This
    commit reverts the changes in an attempt to reclaim that binary memory.
    
    For reference this original commit that converted these
    b551562
Commits on Jan 12, 2015
  1. Alex Arnell
Commits on Jan 8, 2015
  1. Alex Arnell

    suspend the supervisor

    voidlock authored
  2. Alex Arnell
  3. Alex Arnell

    fixup typo in upgrade script

    voidlock authored
  4. Alex Arnell

    fix failing full_buffer_temp_fail test

    voidlock authored
    This test no longer relies on flaky timing and waiting for mocked calls
    to trigger the next series of events. Instead the events are all sent at
    once and the mocked out http server controls when the state flips from a
    500 status to a 200 status.
Commits on Jan 7, 2015
  1. Alex Arnell

    refactored backoff and logging

    voidlock authored
    Removed some unnecessary logging and improved the logging around the
    backoff and buffer resizing.
Commits on Dec 20, 2014
  1. Alex Arnell
  2. Alex Arnell
Commits on Dec 19, 2014
  1. Alex Arnell

    Merge pull request #112 from heroku/http-drains-doc-update

    voidlock authored
    Update drains documentation to reflect reality.
  2. Alex Arnell

    remove logging around code_change

    voidlock authored
    this may help speed the process up a git
Commits on Dec 18, 2014
  1. Alex Arnell

    fix logplex_http_drain code_change

    voidlock authored
    Code changes are hard. Added more logging around the code_change and
    fixed up the return tuple so that it is the expected format.
Commits on Dec 17, 2014
  1. Alex Arnell

    add v78_v79 live_upgrade and code_change code

    voidlock authored
    This change set is isolated to a single module and the live upgrade is
    fairly straight foward. Just need to pause all the http drains, load the
    new module, perform the code_change and resume the drains.
  2. Alex Arnell

    fix failing shrink tests

    voidlock authored
    The shrink test incorrectly asserted the old behaviour of resizing the
    queue as soon as socket is connected. Once the test was fixed I
    discovered and fixed an issue in the logplex_http_drain code that still
    set the service level back to normal as soon as the socket is connected
    instead of when a successfull HTTP request is sent.
  3. Andrew Gwozdziewycz
Commits on Dec 4, 2014
  1. Alex Arnell
  2. Alex Arnell

    make last_good_time more clear

    voidlock authored
    While fixing a failing test I noticed that the drain buffer would cycle
    between the shrunk sizing (10) and the default size (1024) rapidly
    between reconnect attempts. This is an edge case surrounding a drain
    that has never connected before. I altered the last_good_time atom to be
    more clear in it's intension and simplified the maybe_shrink code a
    little.
Commits on Dec 1, 2014
  1. Alex Arnell

    implement retry using exponential backoff

    voidlock authored
    Quick reconnects may help reduce L10 errors when minor networking blips
    are encountered. In the face of longer networking issues, the
    exponential backoff will quickly grow the reconnect delay to enforce
    reasonable rate limits.
  2. Alex Arnell

    save some cycles and hibernate after timeout

    voidlock authored
    In testing I noticed that the idle and max_ttl closing causes back pressure on
    a very busy drain. This change fixed the issue.
Commits on Nov 26, 2014
  1. Edward Muller

    Update Travis

    freeformz authored
Something went wrong with that request. Please try again.