machine-profile annotation design

[ArnobKumarSaha]

Style 1: string

kubernetes.io/instance-type: "t.medium". # db
kubernetes.io/instance-type: "t.small,t.xlarge" # autoscaler

MariaDB
Memcached
MSSQLServer
MySQL
PerconaXtraDB
PgBouncer
Pgpool 
Postgres 
ProxySQL 
RabbitMQ 
Redis 
ZooKeeper 

Style 2: json

kubernetes.io/instance-type: '{"standalone":"t.medium"}' # db
kubernetes.io/instance-type: '{"standalone":"t.small,t.xlarge"}' # autoscaler

kubernetes.io/instance-type: '{"configServer":"t.medium","mongos":"t.small","shard":"t.medium"}'
kubernetes.io/instance-type: '{"configServer":"t.small,t.xlarge","mongos":"t.small,t.xlarge","shard":"t.small,t.xlarge"}'

ClickHouse: 
Druid: brokers, coordinators, historicals, middleManagers
Elasticsearch: combined(node), data, ingest, master
FerretDB: primary, secondary
Kafka: combined(node), broker, controller
MongoDB: standalone, replicaSet, shard, configServer, mongos, arbiter, hidden
Singlestore: standalone(node), aggregator, leaf
Solr: standalone(node), replicaSet(node), coordinator, data, overseer

TODOs:

• singlestore mismatch between db:standlone & autoscaler:node