fix: making ops dashboard work locally + adding runs to database now + other improvements

[0x0elliot]

No description provided.

[0x0elliot]

the last commit was a little brave. it adds the time related apps we have in the original workflow. i removed it a bit ago because it failed a lot. now it doesn't for some reason.

[0x0elliot]

don't merge in yet. i need to add a consistent sleep period between the pings.

[0x0elliot]

now, ops dashboard doesn't feel like a warzone anymore :)

[0x0elliot]

please pull Shuffle/Shuffle#1211 quickly to avoid weird behaviour.

[0x0elliot]

This PR also fixes weird import errors. At least, the ones i see for local instances of shuffle. ones with this in the logs:

{
  "error" : {
    "root_cause" : [
      {
        "type" : "illegal_argument_exception",
        "reason" : "mapper [workflow.actions.position.y] cannot be changed from type [float] to [long]"
      }
    ],
    "type" : "illegal_argument_exception",
    "reason" : "mapper [workflow.actions.position.y] cannot be changed from type [float] to [long]"
  },
  "status" : 400