Cluster API implemenation

gistfile1.txt

Create cluster call

- Define specific stuff for each instance: security group/ruls, dns entry, floating IP etc.
- Connect node group into common network (VLAN, tunneling, flat network)
- Spin-up guestagent on each instance.
- Check statuses of each instance. Generate cluster status.
- Verify by updating cluster status at DВ.

gistfile2.txt

List clusters call

- Collect clusters.
- Lookup over cluster nodes.
- Collect list of IDs
-------------------------------------------------------------
Detailed list cluster call

- Collect clusters.
- Lookup over cluster nodes.
- Retrieve instances information.
- Collect instance specific info.
-------------------------------------------------------------
* info could be defined by passing a dict of instance object attributes
** this could be the same function. Just pass flag detailed=True/False

gistfile3.txt

Drop instance

- Lookup for cluster.
- Lookup for instance.
- Check permissions for minimum nodes.
- If permissions allows to drop, than do. Update status at DB.
- Perform cluster lookup, redefine group of instances. **
-------------------------------------------------------------

* this function could drop more than one instance at the same time, 
  if each drop process will be wrapped into single thread, just another task-to-do

** This mechanism is custom for each service engine

gistfile4.txt

Add instance to cluster/replication

- Spin-up instance. 
- Connect instance to cluster network
- Mark instance with cluster ID.
- Update clusters meta-data with new instance ID.
- Update cluster(next steps are based on implementation) **
- Verify statuses.

-------------------------------------------------------------
** This mechanism is custom for each service engine

gistfile5.txt

Promotion

- If cluster type allows to do that.
- If no master.
- Update cluster.**
- Spin-up cluster.
- Verify statuses.

-------------------------------------------------------------
** This mechanism is custom for each service engine

gistfile6.txt

Restart

- If cluster exists
- If status is not Busy/Error/Building etc.
- If each of nodes are not busy.
- Do restart of each instance.
- Collect statuses.
- Verify cluster status.

gistfile7.txt

Backup cluster

- If cluster exists.
- Discover instances.
- Mark cluster with Busy status.
- Do async backup of each instance.
- Create cluster backup DB entry.
- Verify backup statuses.

gistfile8.txt

Restore cluster from backup

- If backup.
- If backup of each instance.
- If new cluster size is the same as in backup.
- Create each instance from backup.
- Verify statuses.