Meeting 2017 06 29

• Time: 16:15 - 19:30

• Participants: Tim, Matthias, Lennart

• Lennart

  • gRPC failover completed, hopefully reasonable failover configuration values and behaviour
  • Started Maildir Prometheus exporter: maildir_exporter
  • Did not work on IMAP benchmark continuation and pluto exporter to Prometheus, will follow next

• Tim

  • GlusterFS mounted in Dovecot cluster, currently not in "test" usage
  • Moved IMAP benchmark into cluster
  • TODO: Add possibly missing items

• Matthias

  • We will probably not use the Dovecot Prometheus exporter
  • Investigation into logging and Docker issues, in progress, not completed yet
  • Pluto deployed in a new GCE cluster, yeah! (We will fix TLS very soon)
  • First look into GCE federation

• TO

  • Evaluation goals & setup & metrics
    • The whiteboard was filled with lots of interesting ideas

• ToDos until 2017-07-xx

  • Matthias:
    • Have a look how to integrate a failure framework into our clusters, see Evaluation
    • gokit level.go:63 default caller problem
    • Fail of make docker build pipeline step
    • Look into GCE federation so that we can make use of it
    • Together with Lennart, fix TLS bad certificate error in internal communication
    • TODO
  • Lennart:
    • Move all repos of pluto on GitHub into a GitHub organization
    • Together with Matze, fix TLS bad certificate error in internal communication
    • Have a look into how we could handle the too many open file descriptors situation
    • Add cancellation of gRPC background connection routines that are long gone
    • Finish Maildir exporter
    • Meet with Tim to push evaluation of pluto and discuss first two evaluation cases
    • gokit level.go:63 default caller problem
  • Tim:
    • TODO