1. 21 Aug, 2020 10 commits
  2. 20 Aug, 2020 11 commits
  3. 19 Aug, 2020 4 commits
    • aguestuser's avatar
      Merge branch 'hotfix-register-healthcheck-callback-before-sending' into 'master' · b4b2e231
      aguestuser authored
      [hotfix] register healtcheck callback before sending
      
      See merge request team-friendo/signalboost!353
      b4b2e231
    • aguestuser's avatar
      [hotfix] register healtcheck callback before sending · e957bdc8
      aguestuser authored
      context:
      * after raising socket pool size from 1 to 5 we observed a dramatic
        spike in false negative healthchecks
      * it is still unclear why, but one possible explanation is that
        because messages are being sent faster, we are sending healthchecks
        before we have a chance to register their callback, thus the
        healthcheck response arrives before its callback handler has been
        registered, and the callback handler, once registered, waits for a
        healthcheck response that never arrives (because it arrived before
        the handler started listening)
      
      changes:
      * this MR tests the above hypothesis by registering healtcheck
        callbacks before sending them. if it works, we should observe a
        decrease to close to zero false negative healthchecks
      e957bdc8
    • aguestuser's avatar
      Merge branch 'hotfix-bump-socket-pool-to-5' into 'master' · 9a665439
      aguestuser authored
      [hotfix] bump socket pool to 5
      
      See merge request team-friendo/signalboost!352
      9a665439
    • aguestuser's avatar
      [hotfix] bump socket pool to 5 · fedb94c1
      aguestuser authored
      context:
      * as a test to see if it would result in fewer concurrency bugs in
        signald, we dropped the signalboost socket pool size from 50 to 1
      * over a 5 day period, we observed only 1 channel crash (fixable by
        restarting), but we saw average message delivery latency jump from
        around 50ms-1sec to 30-60sec.
      * we also saw ~10 false positive failed health checks per day
      
      change:
      * as an experiement to see if we can bring delivery latency down
        (and reduce false positives) without drastically increasing crashes,
        we are going to raise the socket pool size from 1 to 5 and observe
        the effects for 5 days
      fedb94c1
  4. 18 Aug, 2020 5 commits
  5. 17 Aug, 2020 10 commits