Commits on May 6, 2013
  1. @archaelus
  2. @archaelus

    Have nsync cache tokens.

    archaelus committed May 6, 2013
  3. @archaelus
Commits on May 4, 2013
  1. @archaelus
  2. @archaelus

    Cache tokens in #channel records.

    Adds logplex_channel:
     - tokens(Channel)
     - has_token(Id, Channel)
     - cache(Channel)
     - cache_token(ChannelId, Token)
    archaelus committed May 3, 2013
  3. @archaelus
  4. @archaelus
  5. @archaelus

    Add logplex_token:cache/1.

    archaelus committed May 3, 2013
  6. @archaelus
  7. @archaelus
  8. @archaelus
Commits on May 3, 2013
  1. @ferd

    Removing lengthy redis checks

    Looking for drains is too expensive as a regular operation. In
    production it may take up to two minutes just to run all the checks and
    Redis is what makes it slow. Tests have shown that no drain currently on
    a production node appears to be orphaned in either ETS or Redis, but
    always both. As such, we drop this lookup.
    ferd committed May 3, 2013
Commits on May 2, 2013
  1. @ferd

    Reducing delay to 5 minutes for leak monitor

    The operations are long-lived and there isn't a lot of expected benefits
    to having things run very frequently. A 5 minutes check ought to be
    sufficient.
    ferd committed May 2, 2013
Commits on May 1, 2013
  1. @ferd

    Upgrade path v65 -> v66

    ferd committed May 1, 2013
  2. @ferd

    Increasing kill time on health monitor

    The health monitor's own supervisor gives 8 seconds and it's useless to
    give only 2 to its parent. The 8 seconds might be required because
    scanning entire tables for orphans might be a bit long before seeing
    sys messages.
    ferd committed May 1, 2013
  3. @ferd

    Minor fixes

    ferd committed May 1, 2013
  4. @ferd

    Fixing channel deletion to include drains

    Channel deletion would delete only the ETS table entries for each drain,
    leaving the door open for new nodes and NSYNC'ing to bring back dead
    drains (that were not killed on existing nodes).
    
    This patch makes it so that deleting a channel, which calls
    logplex_drain:delete_by_channel/1 actually removes said drain via redis.
    nsync callbacks will then clean up the rest.
    
    Conflicts:
    	src/logplex_drain.erl
    	src/logplex_mon_serv.erl
    ferd committed Apr 30, 2013
  5. @ferd

    Adding basic monitoring of drains & channels

    Looks for orphans (both in ETS vs. drain, or channels vs. drains, in any
    storage available).
    
    Also looks at total drain and channel count, and compares it with other
    nodes.
    ferd committed Apr 30, 2013
Commits on Apr 22, 2013
  1. @archaelus
  2. @archaelus
  3. @archaelus
Commits on Apr 19, 2013
  1. @archaelus
  2. @archaelus
  3. @archaelus
  4. @archaelus

    Fix malfomed_messages typo.

    archaelus committed Apr 19, 2013
  5. @archaelus
  6. @archaelus
  7. @archaelus
  8. @archaelus
Commits on Apr 18, 2013
  1. @archaelus

    Simplify process_post with respond/4.

    Also provides better error messages.
    archaelus committed Apr 17, 2013
  2. @archaelus
  3. @archaelus
  4. @archaelus
Commits on Apr 17, 2013
  1. @archaelus

    Merge branch 'v65'

    archaelus committed Apr 17, 2013
  2. @archaelus