CPU measurement tags are not returned

Tags are not loading at all for “CPU” measurement.. for other measurements its loading properly.

Upon investigating , found that Chronograf posts following metadata query to InfluxDb when we select CPU metric in chronograf UI and it goes to pending state as there is no response back from InfluxDB for longer duration and then times out

`SHOW TAG VALUES FROM "autogen"."cpu" WITH KEY IN ("GlobalMon.Site", "__auth_key", "_resourcegroup_", "_subscription_", "_team_", "application", "cluster_owner", "cpu", "cynosure.isCynosure", "dc", "dspa.xdm application", "dspa.xdm environment", "env", "host", "hostname", "logfile", "month", "namespace", "weekday", "year")"`

This metadata query works fine when posted without Tag key "CPU".

`SHOW TAG VALUES FROM "autogen"."cpu" WITH KEY IN ("GlobalMon.Site", "__auth_key", "_resourcegroup_", "_subscription_", "_team_", "application", "cluster_owner", "cynosure.isCynosure", "dc", "dspa.xdm application", "dspa.xdm environment", "env", "host", "hostname", "logfile", "month", "namespace", "weekday", "year")"`

TAG value cardinality for key "CPU "- in CPU measurement is 418

Tried posting following query manually in Influx data node command line and it takes longer duration and never loads.

`SHOW TAG VALUES ON "vc" FROM "autogen"."cpu" WITH KEY IN ("cpu")`

Any pointers how to troubleshoot this ?

Start by reproducing the Chronograf CPU-metric metadata request and the manual `SHOW TAG VALUES ON "vc" FROM "autogen"."cpu" WITH KEY IN ("cpu")` query in InfluxDB. Trace why the CPU tag query remains pending, then verify that CPU measurement tags return within a usable time and no longer time out.

Written by the indexing model from the issue text.