TEST-PLAN.md

Cassandra release test plan Unit tests Templates to tests

Included:

cassandre.yaml
post-start.sh

Rendering technologies:

Test templates with bosh-template 2.0.0 Gem

Configurations

Mandatory
    replication mono datacenter
    replication multi-datacenter (si possible)

Optional
    SSL or not
    Rac Aware or not

Acceptance tests

Write BOSH errand job. Common tests

These test will be run for both un seul noeud and avec Replication

Read/Write test

Before each:
    connect as admin
    create a keyspace
    create user with admin or without role on the database ('SUPERUSER')

Test case:
    connect as user
    create table
    write some data in the table
    read the data back
    update the data
    delete data
    verify data is absent

After each:
    drop keyspace
    drop user

Backup/Restore

Test case:
    write some data
    cassandra backup
    drop keyspace
    restore keyspace
    read the data back

Replication factor (RF)= 3
    Verify that the data is well replicated
      by inserting some data into cassandra and verified that the correspoding token are also replicated on other nodes (2 if RF=3) 

Soft Failover
    Build a 3-nodes cassandra : 1 seed and 2 servers
    Turn the resurrector off
    Stop on off the server node gracefully (stopdaemon)
    Verify that you still have the same result for the same CQL request ('select count(*) from keyspace_name.table_name')

Cluster re-join after graceful stop

    Same begining as above
    Start the node back 
    verify it joins the cluster in Up and Running mode
    verify the data is here

Dirty Failover
    Build a 3-nodes cassandra : 2 seeds and 2 servers
    Turn the resurrector on
    Kill one seed node 
    Verify that the last seed node takes and the rest of the cassandra cluster is always available

Cluster recovery

    Same begining as above
    Wait for the resurrector to recreate the killed node
    verify

Scale-in (general case)
    Build 3-nodes cassandra : 1 seed and 2 servers
    Scale in to 1 server node (decommission and removenode)
    Verify that deleted nodes have properly left the cluster

Scale-in (involing a seed node)
    Build 3-nodes cassandra : 2 seed and 2 servers
    Scale in to 1 seed node (decommission and remove node)
    Verify that deleted seed node have properly left the cluster nd the rest of the cassandra cluster is always available

Scale-out
    Build 3-nodes cassandra : 1 seed and 2 servers
    Scale out to 3 servers nodes
    Verify that new nodes join the cluster properly

SSL
    Verify SSL is properly implemented



ALTER USER
CREATE USER
DROP USER
LIST USERS