Spring Security With Cassandra-Authentication Failed - spring-security

Question

I'm Trying to authenticate user using spring security with Cassandra database. I'm getting following exception:
Authentication Failed: PreparedStatementCallback; bad SQL grammar [select userName, password from user where userName=?];
nested exception is java.sql.SQLSyntaxErrorException:
No indexed columns present in by-columns clause with "equals" operator 'select userName, password from user where userName=?'
Here is configuration details:
security-config.xml
<form-login
login-processing-url="/j_spring_security_check"
login-page="/index.xhtml" always-use-default-target="true"
authentication-failure-url="/index.xhtml?error=true"
authentication-success-handler-ref="authenticationSuccessHandler"
authentication-failure-handler-ref="authenticationFailureHandler"/>
<authentication-manager alias="authenticationManager">
<authentication-provider>
<jdbc-user-service data-source-ref="cassandraDataSource" users-by-username-query="select userName, password from user where userName=?"
authorities-by-username-query="select roleId, userName, roleName from role where userName=?"/>
</authentication-provider>
</authentication-manager>
and I created the index with cql command:
CREATE INDEX userName_index ON role (userName);
CREATE INDEX userNameUser_index ON user (userName);
result of desc mykeyspaces
CREATE TABLE role (
roleid int PRIMARY KEY,
rolename text,
username text
) WITH bloom_filter_fp_chance = 0.01
AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
AND comment = ''
AND compaction = {'min_threshold': '4', 'class': 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy', 'max_threshold': '32'}
AND compression = {'sstable_compression': 'org.apache.cassandra.io.compress.LZ4Compressor'}
AND dclocal_read_repair_chance = 0.1
AND default_time_to_live = 0
AND gc_grace_seconds = 864000
AND max_index_interval = 2048
AND memtable_flush_period_in_ms = 0
AND min_index_interval = 128
AND read_repair_chance = 0.0
AND speculative_retry = '99.0PERCENTILE';
CREATE INDEX userName_index ON role (username);
CREATE TABLE user (
userid int PRIMARY KEY,
email text,
firstname text,
lastname text,
password text,
phone int,
username text
) WITH bloom_filter_fp_chance = 0.01
AND caching = '{"keys":"ALL", "rows_per_partition":"NONE"}'
AND comment = ''
AND compaction = {'min_threshold': '4', 'class': 'org.apache.cassandra.db.compaction.SizeTieredCompactionStrategy', 'max_threshold': '32'}
AND compression = {'sstable_compression': 'org.apache.cassandra.io.compress.LZ4Compressor'}
AND dclocal_read_repair_chance = 0.1
AND default_time_to_live = 0
AND gc_grace_seconds = 864000
AND max_index_interval = 2048
AND memtable_flush_period_in_ms = 0
AND min_index_interval = 128
AND read_repair_chance = 0.0
AND speculative_retry = '99.0PERCENTILE';
CREATE INDEX userNameUser_index ON user (username);
cqlsh:system>

Answer 1

If you are using user and role tables for security, username should be a partition key on both of those tables. Just indexing them is going to cause querying multiple nodes and not going to be effective. The data model below will be more effective for your queries.
CREATE TABLE user (
userid int,
email text,
firstname text,
lastname text,
password text,
phone int,
username text,
PRIMARY KEY ((username), userid)
)
CREATE TABLE role (
roleid int,
rolename text,
username text,
PRIMARY KEY ((username), roleid)
)
As you can see, primary key consists of 2 columns, to insure uniqueness, but only the first one is a partition key. This insures that all records for that user will be on one node and queries by user name will be effective.
There is a lot of info available about data modeling for Cassandra which differs a lot from data modeling for relational database. The basic rules are:
Understand partition and cluster keys
Model to your queries, not data
Indexes are not efficient