A sample health check script to get 1 row from various partitions of the tables in a YCQL keyspace

ycql_table_health_v1.py

# pip install yb-cassandra-driver

from cassandra.cluster import Cluster

# cluster = Cluster(['127.0.0.1'])
cluster = Cluster(['172.151.30.71', '172.151.28.193'])

num_checks_per_table=64
keyspace_name="ybdemo_keyspace"

session = cluster.connect()

def check_table_health(keyspace_name, table_name):
    print("Checking health for: " + keyspace_name + "." + table_name);
    results = session.execute(("SELECT column_name, position  " +
                               "FROM system_schema.columns " +
                               "WHERE keyspace_name = %s AND table_name = %s " +
                               """AND kind='partition_key' """),
                               (keyspace_name, table_name))

    # Add the partition columns to an array sorted by the position of
    # the column in the primary key.
    partition_columns = [''] * 256
    num_partition_columns = 0
    for row in results:
        partition_columns[row.position] = row.column_name
        num_partition_columns = num_partition_columns + 1
    del partition_columns[num_partition_columns:] # remove extra null elements from array

    p_columns = ",".join(partition_columns)
    print("Partition columns for " + keyspace_name + "." + table_name + ": (" + p_columns + ")");

    stmt = ("SELECT {} FROM {}.{} " +
            "WHERE partition_hash({}) >= ? " +
            "AND partition_hash({}) <= ? LIMIT 1").format(p_columns,
                                                           keyspace_name,
                                                           table_name,
                                                           p_columns,
                                                           p_columns)
    print("Stmt = {}".format(stmt))

    prepared_stmt = session.prepare(stmt)
    range_size = (64*1024)/num_checks_per_table
    print("Performing {} checks for {}.{}".format(num_checks_per_table, keyspace_name, table_name))
    for idx in range(num_checks_per_table):
        l_bound = idx * range_size
        u_bound = l_bound + range_size - 1
        print("Checking {} to {} partition...".format(l_bound, u_bound))
        # This may return 0 or 1 rows (because of LIMIT 1), but we are not really interested
        # in the result. This is just a check to see if the tablet backing the query is alive
        # (has a leader) and is responding.
        results = session.execute(prepared_stmt, (l_bound, u_bound))
    print("--------------------------------------------------------")

def check_keyspace_health(keyspace_name):
    print("Checking health for keyspace: " + keyspace_name);
    print("--------------------------------------------------------")
    tables = []
    results = session.execute("select table_name from system_schema.tables where keyspace_name = %s",
                              (keyspace_name, ));
    for row in results:
       check_table_health(keyspace_name, row.table_name)

# Main
check_keyspace_health(keyspace_name)