[cassandra_check] Add cassandra_check

.travis.yml

@@ -42,6 +42,9 @@ env:
     - TRAVIS_FLAVOR=apache FLAVOR_VERSION=2.4.12
     - TRAVIS_FLAVOR=cassandra FLAVOR_VERSION=2.0.17
     - TRAVIS_FLAVOR=cassandra FLAVOR_VERSION=2.1.14
+    - TRAVIS_FLAVOR=cassandra_check FLAVOR_VERSION=2.0.17
+    - TRAVIS_FLAVOR=cassandra_check FLAVOR_VERSION=2.1.14
+    - TRAVIS_FLAVOR=cassandra_check FLAVOR_VERSION=3.10
     - TRAVIS_FLAVOR=couch FLAVOR_VERSION=1.6.1
     - TRAVIS_FLAVOR=consul FLAVOR_VERSION=v0.6.4
     - TRAVIS_FLAVOR=consul FLAVOR_VERSION=0.7.2

cassandra_check/CHANGELOG.md

@@ -0,0 +1,8 @@
+# CHANGELOG - Cassandra_check
+
+0.1.0/ Unreleased
+==================
+
+### Changes
+
+* [FEATURE] adds cassandra_check integration.

cassandra_check/README.md

@@ -0,0 +1,29 @@
+# Cassandra Check
+
+## Overview
+
+Get metrics from cassandra databases that are not available through the [jmx integration](https://github.com/DataDog/integrations-core/tree/master/cassandra)
+
+## Installation
+
+Install the `dd-check-cassandra_check` package manually or with your favorite configuration manager
+
+## Configuration
+
+Edit the `cassandra_check.yaml` file to point to your server and port and set the keyspaces to monitor
+
+## Validation
+
+When you run `datadog-agent info` you should see something like the following:
+
+    Checks
+    ======
+
+        cassandra_check
+        -----------
+          - instance #0 [OK]
+          - Collected 39 metrics, 0 events & 7 service checks
+
+## Compatibility
+
+The cassandra_check check is compatible with all major platforms

cassandra_check/check.py

@@ -0,0 +1,58 @@
+# (C) Datadog, Inc. 2010-2016
+# All rights reserved
+# Licensed under Simplified BSD License (see LICENSE)
+
+# 3rd party
+from cassandra.cluster import Cluster, NoHostAvailable
+from cassandra.auth import PlainTextAuthProvider
+
+# project
+from checks import AgentCheck
+
+EVENT_TYPE = SOURCE_TYPE_NAME = 'cassandra_check'
+DEFAULT_NODE_IP = 'localhost'
+DEFAULT_NODE_PORT = 9042
+
+
+class CassandraCheck(AgentCheck):
+
+    def __init__(self, name, init_config, agentConfig, instances=None):
+        AgentCheck.__init__(self, name, init_config, agentConfig, instances)
+
+    def check(self, instance):
+        # Get the node IP address to connect Cassandra
+        node_ip = instance.get("node_ip", DEFAULT_NODE_IP)
+        node_port = instance.get("node_port", DEFAULT_NODE_PORT)
+        keyspaces = instance.get("keyspaces", [])
+        tags = instance.get("tags", [])
+        connect_timeout = instance.get("connect_timeout", 5)
+
+        username = instance.get("username", "")
+        password = instance.get("password", "")
+        auth_provider = PlainTextAuthProvider(username, password)
+
+        # Try to connect to the node
+        cluster = Cluster([node_ip], port=node_port, auth_provider=auth_provider, connect_timeout=connect_timeout)
+        try:
+            cluster.connect(wait_for_all_pools=True)
+            if keyspaces:
+                for keyspace in keyspaces:
+                    token_map = cluster.metadata.token_map
+                    down_replicas = 0
+                    for token in token_map.ring:
+                        replicas = token_map.get_replicas(keyspace, token)
+                        down_replicas = max(down_replicas, len([r for r in replicas if not r.is_up]))
+
+                    self.gauge("cassandra.replication_failures", down_replicas,
+                        tags=["keyspace:%s" % keyspace, "cluster:%s" % cluster.metadata.cluster_name] + tags)
+
+
+        except NoHostAvailable as e:
+            self.log.error('Could not connect to node %s:%s : %s' % (node_ip, node_port, e))
+            node_status = AgentCheck.CRITICAL
+        else:
+            node_status = AgentCheck.OK
+        finally:
+            cluster.shutdown()
+
+        self.service_check('cassandra.can_connect', node_status, tags=tags)

cassandra_check/ci/cassandra_check.rake

@@ -0,0 +1,99 @@
+require 'ci/common'
+
+def cassandra_check_version
+  ENV['FLAVOR_VERSION'] || '2.1.14' # '2.0.17'
+end
+
+def cassandra_check_rootdir
+  "#{ENV['INTEGRATIONS_DIR']}/cassandra_check_#{cassandra_check_version}"
+end
+
+container_name = 'dd-test-cassandra'
+container_name2 = 'dd-test-cassandra2'
+
+namespace :ci do
+  namespace :cassandra_check do |flavor|
+    task before_install: ['ci:common:before_install'] do
+      sh %(docker kill #{container_name} 2>/dev/null || true)
+      sh %(docker rm #{container_name} 2>/dev/null || true)
+      sh %(docker kill #{container_name2} 2>/dev/null || true)
+      sh %(docker rm #{container_name2} 2>/dev/null || true)
+    end
+
+    task :install do
+      Rake::Task['ci:common:install'].invoke('cassandra_check')
+      sh %(docker create --expose 9042 --expose 7000 --expose 7001 --expose 9160 \
+            -p 9042:9042 -p 7000:7000 -p 7001:7001 -p 9160:9160 --name #{container_name} cassandra:#{cassandra_check_version})
+      sh %(docker start #{container_name})
+      sh %(docker create --name #{container_name2} \
+            -e CASSANDRA_SEEDS="$(docker inspect --format='{{ .NetworkSettings.IPAddress }}' #{container_name})" cassandra:#{cassandra_check_version})
+      sh %(docker start #{container_name2})
+    end
+
+    task before_script: ['ci:common:before_script'] do
+      # Wait.for container_port
+      count = 0
+      logs = `docker logs #{container_name} 2>&1`
+      logs2 = `docker logs #{container_name2} 2>&1`
+      puts 'Waiting for Cassandra to come up'
+      until count == 20 || ((logs.include?('Listening for thrift clients') || logs.include?('Starting listening for CQL clients')) && \
+             (logs2.include?('Listening for thrift clients') || logs2.include?('Starting listening for CQL clients')))
+        sleep_for 4
+        logs = `docker logs #{container_name} 2>&1`
+        logs2 = `docker logs #{container_name2} 2>&1`
+        count += 1
+      end
+      if (logs.include?('Listening for thrift clients') || logs.include?('Starting listening for CQL clients')) && \
+         (logs2.include?('Listening for thrift clients') || logs2.include?('Starting listening for CQL clients'))
+        puts 'Cassandra is up!'
+      else
+        puts 'Logs of container 1'
+        sh %(docker logs #{container_name} 2>&1)
+        puts 'Logs of container 2'
+        sh %(docker logs #{container_name2} 2>&1)
+        raise
+      end
+    end
+
+    task script: ['ci:common:script'] do
+      this_provides = [
+        'cassandra_check'
+      ]
+      Rake::Task['ci:common:run_tests'].invoke(this_provides)
+    end
+
+    task before_cache: ['ci:common:before_cache']
+
+    task cleanup: ['ci:common:cleanup'] do
+      sh %(docker kill #{container_name} 2>/dev/null || true)
+      sh %(docker rm #{container_name} 2>/dev/null || true)
+      sh %(docker kill #{container_name2} 2>/dev/null || true)
+      sh %(docker rm #{container_name2} 2>/dev/null || true)
+    end
+
+    task :execute do
+      exception = nil
+      begin
+        %w(before_install install before_script).each do |u|
+          Rake::Task["#{flavor.scope.path}:#{u}"].invoke
+        end
+        if !ENV['SKIP_TEST']
+          Rake::Task["#{flavor.scope.path}:script"].invoke
+        else
+          puts 'Skipping tests'.yellow
+        end
+        Rake::Task["#{flavor.scope.path}:before_cache"].invoke
+      rescue => e
+        exception = e
+        puts "Failed task: #{e.class} #{e.message}".red
+      end
+      if ENV['SKIP_CLEANUP']
+        puts 'Skipping cleanup, disposable environments are great'.yellow
+      else
+        puts 'Cleaning up'
+        Rake::Task["#{flavor.scope.path}:cleanup"].invoke
+      end
+      raise exception if exception
+    end
+  end
+end

cassandra_check/conf.yaml.example

@@ -0,0 +1,15 @@
+init_config:
+
+instances:
+  # Configuration options:
+  # keyspaces: a list of keyspaces to monitor
+  # node_ip: the IP of the cassandra node to connect to. The rest of the nodes in the cluster will be auto discovered by the cassandra driver 
+  #   For more information, see https://datastax.github.io/python-driver/api/index.html
+  #   Default to localhost.
+  # node_port: the port cassandra is listening for connections.
+  #   Default to 9042
+  # username/password: a set of credentials to connect to the cassandra cluster.
+  # tags: optional, a list of tags to be sent with the metrics
+  # connect_timeout: timeout, in seconds, for creating new connections. Default to 5. 
+
+  - keyspaces: ["foo"]

cassandra_check/manifest.json

@@ -0,0 +1,12 @@
+{
+  "maintainer": "help@datadoghq.com",
+  "manifest_version": "0.1.0",
+  "max_agent_version": "6.0.0",
+  "min_agent_version": "5.6.3",
+  "name": "cassandra_check",
+  "short_description": "cassandra_check description.",
+  "guid": "00e4a8bd-8ec2-4bb4-b725-6aaa91618d13",
+  "support": "contrib",
+  "supported_os": ["linux","mac_os","windows"],
+  "version": "0.1.0"
+}

cassandra_check/metadata.csv

@@ -0,0 +1,2 @@
+metric_name,metric_type,interval,unit_name,per_unit_name,description,orientation,integration,short_name
+cassandra.replication_failures,gauge,,,,Number of replica nodes down per keyspace,-1,cassandra_check,replication failures

cassandra_check/requirements.txt

@@ -0,0 +1,2 @@
+# integration pip requirements
+cassandra-driver==3.10.0

cassandra_check/test_cassandra_check.py

@@ -0,0 +1,43 @@
+# (C) Datadog, Inc. 2010-2016
+# All rights reserved
+# Licensed under Simplified BSD License (see LICENSE)
+
+# stdlib
+from nose.plugins.attrib import attr
+
+# 3p
+from cassandra.cluster import Cluster
+
+# project
+from tests.checks.common import AgentCheckTest
+
+
+@attr(requires='cassandra_check')
+class TestCassandraCheck(AgentCheckTest):
+    """Basic Test for cassandra_check integration."""
+    CHECK_NAME = 'cassandra_check'
+
+    def test_check(self):
+
+        # Create a keyspace with replication factor 2
+        cluster = Cluster(connect_timeout=1)
+        session = cluster.connect()
+        session.execute("CREATE KEYSPACE IF NOT EXISTS test WITH REPLICATION = {'class': 'SimpleStrategy', 'replication_factor': 2}")
+        cluster.shutdown()
+
+        config = {
+            'instances': [
+                {'host': '127.0.0.1',
+                'port': 9042,
+                'keyspaces': ['test'],
+                'tags': ['foo','bar'],
+                'connect_timeout': 1}
+            ]
+        }
+
+        self.run_check(config)
+        # We should have the value 1 since the driver won't be able to connect to one of the container (port not exposed)
+        self.assertMetric('cassandra.replication_failures', value=1, tags=['keyspace:test', 'cluster:Test Cluster', 'foo', 'bar'])
+        self.assertServiceCheckOK('cassandra.can_connect', tags=['foo', 'bar'])
+        # Raises when COVERAGE=true and coverage < 100%
+        self.coverage_report()

circle.yml

@@ -80,6 +80,7 @@ test:
         - rake ci:run[kafka]
         - rake ci:run[docker_daemon]
         - rake ci:run[kubernetes]
+        - rake ci:run[cassandra_check]
         - bundle exec rake requirements
     post:
         - if [[ $(docker ps -a -q) ]]; then docker stop $(docker ps -a -q); fi