All Sahara Cluster operations are performed in multiple steps. A Cluster object has a Status attribute which changes when Sahara finishes one step of operations and starts another one.
Before performing any operations with OpenStack environment, Sahara validates user input.
If any of validations fails, the Cluster will still be kept in database with Error status.
This status means that the Provisioning plugin performs some infrastructural updates.
It takes some time for OpenStack to schedule all required VMs and Volumes, so Sahara wait until all of them are in Active state.
Sahara waits while VMs’ operating systems boot up and all internal infrastructure components like networks and volumes are attached and ready to use.
Sahara preparers a Cluster for starting. This step includes generating /etc/hosts file, so that all instances could access each other by a hostname. Also Sahara updates authorized_keys file on each VM, so that communications could be done without passwords.
Sahara pushes service configurations to VMs. Both XML based configurations and environmental variables are set on this step.
Sahara is starting Hadoop services on Cluster’s VMs.
Active status means that a Cluster has started successfully and is ready to run Jobs.
Sahara checks the scale/shrink request for validity. The Plugin method called for performing Plugin specific checks is different from creation validation method.
Sahara performs database operations updating all affected existing Node Groups and creating new ones.
State similar to Spawning while Custer creation. Sahara adds required amount of VMs to existing Node Groups and creates new Node Groups.
State similar to Configuring while Cluster creation. New instances are being configured in the same manner as already existing ones. Existing Cluster VMs are also updated with a new /etc/hosts file.
Sahara stops Hadoop services on VMs that will be deleted from a Cluster. Decommissioning Data Node may take some time because Hadoop rearranges data replicas around the Cluster, so that no data will be lost after tht VM is deleted.
The same Active as after Cluster creation.
The only step, that releases all Cluster’s resources and removes it form database.
If Cluster creation fails, the Cluster will get into Error state. This state means the Cluster may not be able to perform any operations normally. This cluster will stay in database until it is manually deleted. The reason of failure may be found in Sahara logs.
If an error occurs during Adding Instances operation, Sahara will first try to rollback this operation. If rollback is impossible or fails itself, then the Cluster will also get into Error state.