New Advanced query settings in cluster

The following cluster level Query setting is required (UI, ns_server,query support)

Query parameter name: txtimeout
Default values : "0ms"

Query parameter name: n1ql-feat-ctrl
Current default value : 12
New default value : 76

parameter name : memory-quota
Default values : 0
Quantity : MB

parameter name : use-cbo
Default value : true

 For the below parameters, we want it to be set at the global level and hence in ns_server. We do not want to expose this to the end-user via UI yet.  

parameter name : cleanupClientAttempts
Default values :true
Purpose: Controls whether the main cleanup thread is running, which cleans up any attempts created by this process.

parameter name : cleanupLostAttempts
Default values :true
Purpose: Controls whether the lost/polling cleanup thread is running, which cleans up expired attempts from any process.  By default it checks the default collection of all buckets, but this can be controlled with next tunable.

parameter name : cleanupWindow
Default values :60s
Purpose: Controls how long the polling cleanup takes to check all ATRs.

parameter name : numAtrs
Default values : 1024
Purpose: There is a list of 20k ATRs that may be used.  This is an experimental parameter that gives us an 'out' if people start hitting contention issues with them.
Controls a) pool of ATRs to select from when creating the ATR entry and b) what ATRs should be polled by cleanup.