Database Design

We have several collections, as follows.

Host

Track the information of a Host.

A typical host may look like:

id	name	worker_api	create_ts	capacity	status	clusters	type	log_level	log_type	log_server	autofill	schedulable
xxx	host_0	tcp://10.0.0.1:2375	20160430101010	20	active	[c1,c2,c3]	single	debug	syslog	udp://10.0.0.2:5000	true	true

id (str): uuid of the host instance
name (str): human-readable name
worker_api (str): Through which url to access the Docker/Swarm Daemon
create_ts (datetime): When to add the host
capacity (int): Maximum number of chains on that host
status (str): 'active' (Can access daemon service) or 'inactive' (disconnected from daemon service)
clusters (list): List of the ids of those chains on that host
type (str): 'singe' (single Docker host) or 'swarm' (Docker Swarm cluster)
log_level (str): logging level for chains on the host, e.g., 'debug', 'info', 'warn', 'error'
log_type (str): logging type for chains on the host, 'local' or 'syslog'
log_server (str): log server address, only valid when log_type is 'syslog'
autofill (str): whether to autofill the server to its capacity with chains, 'true' or 'false'
schedulable (str): whether to schedule a chain request to that host, 'true' or 'false', useful when maintain the host

Track information of one blockchain.

A typical cluster may look like:

id	service_url	name	user_id	host_id	worker_api	consensus_plugin	consensus_mode	create_ts	apply_ts	release_ts	duration	size	containers	health
xxx	{}	cluster_A	""	host_xx	tcp://10.0.0.1:2375	pbft	batch	20160430101010	20160430101010			4	[vp0,vp1,vp2,vp3]	OK

id (str): uuid of the host instance
service_url (dict): urls to access the services on the chain, e.g., {'rest':10.0.0.1:7050, 'grpc':10.0.0.1:7051}
name (str): human-readable name
user_id (str): Which user occupies this chain, empty for no occupation
host_id (str): Where the chain exists
worker_api (str): Through which url to access the Docker/Swarm Daemon
consensus_plugin (str): Consensus plugin name
consensus_mode (str): Consensus plugin mode name
create_ts (datetime): When to create the chain
apply_ts (datetime): When the chain is applied
release_ts (datetime): When to release the chain
duration (str): How long the chain lives
size (int): Peer nodes number of the chain
containers (list): List of the ids of those containers for the chain
health (str): 'OK' (healthy status) or 'Fail' (Not healthy)