Skip to content

Kafka and Zookeeper

Preferring local clusters in Frontend and Consumers

To ensure the lowest possible response times, it is recommended to connect to local zookeeper and kafka clusters. In the configuration, you can specify the properties of multiple clusters. For example:

consumer:
  datacenter:
    name:
      source: "ENV"
      env: "DC"
  zookeeper:
    clusters:
      - datacenter: "dc-1"
        connectionString: "zk-1:2181"
      - datacenter: "dc-2"
        connectionString: "zk-2:2181"
  kafka:
    clusters:
      - datacenter: "dc-1"
        brokerList: "kafka-1:9092"
      - datacenter: "dc-2"
        brokerList: "kafka-2:9092"

If you don't specify the {modulePrefix}.datacenter.name.source property hermes will look for a cluster in properties with the datacenter: "dc" property. Hermes also supports retrieving information about the name of the datacenter based on an environment variable. All you have to do is specify {modulePrefix}.datacenter.name.source to ENV just like above, and the name of variable {modulePrefix}.datacenter.name.env in which the datacenter name is stored.

Zookeeper

Hermes uses Zookeeper as metadata store. It does not have to be the same Zookeeper as the one used by Kafka.

Option in Frontend/Consumers Option in Management Description Default value
{modulePrefix}.zookeeper.clusters.[n].connectionString storage.connectionString Zookeeper connection string localhost:2181
{modulePrefix}.zookeeper.clusters.[n].connectionTimeout storage.connectTimeout connection timeout in seconds 10s
{modulePrefix}.zookeeper.clusters.[n].maxRetries storage.retryTimes retry count when connection fails 2
{modulePrefix}.zookeeper.clusters.[n].baseSleepTime storage.retrySleep time to wait between subsequent retries in seconds 1s
{modulePrefix}.zookeeper.clusters.[n].root storage.pathPrefix prefix for Hermes data (if not specified in connection string) /hermes
{modulePrefix}.zookeeper.clusters.[n].authorization.enabled n/a enable Zookeeper authorization false
{modulePrefix}.zookeeper.clusters.[n].authorization.scheme storage.authorization.scheme authorization scheme digest
{modulePrefix}.zookeeper.clusters.[n].authorization.user storage.authorization.user username user
{modulePrefix}.zookeeper.clusters.[n].authorization.password storage.authorization.password password password

Kafka

Single Kafka cluster

In simple case, Hermes is connected to just one Kafka cluster. Frontend and Consumers connect to Kafka to publish and pull messages. Management connects to Kafka to manage existing topics and initiate retransmissions.

Frontend and Consumers options:

Option Description Default value
kafka.broker.list list of all brokers in the cluster (or at least some contact points); separated with ',' localhost:9092
kafka.namespace namespace is a prefix prepended to all Kafka topics and consumer groups used by Hermes
kafka.zookeeper.connect.string [Consumers only] connection string to Kafka Zookeeper localhost:2181
kafka.cluster.name name of Kafka cluster (relevant only when connecting to multiple clusters) primary

Zookeeper connection specific options (retries etc) are read from Metadata Zookeeper options.

Management module can connect to multiple Kafka clusters at once (see section below), thus when specifying connection option is done per cluster. Simple configuration for single cluster looks following:

kafka:
  defaultNamespace: // namespace shared by all clusters, default: <empty>
  clusters:
    -
      clusterName: // name of cluster, can be any arbitrary string, default: primary
      connectionString: // connection string to cluster Zookeeper, default: localhost:2181

Multiple Kafka and Zookeeper clusters

Hermes can be configured to publish and read messages to/from multiple Kafka clusters and to store metadata in multiple Zookeeper clusters. We use this feature on production environment where we have separated kafka clusters in different data centers. If Kafka in one DC fails, whole traffic can be routed to the second DC. This scenario assumes, that Kafka clusters hold different set of messages. There is no support for multiple clusters each holding the same copy of data.

This is the schematics of two data center architecture:

Multi DC schematics

  • there are specific Frontend and Consumers instances per cluster:
  • each Frontend instance writes data to single cluster
  • each Consumers instance reads data from single cluster
  • each Management instance:
  • connects to all Kafka clusters and keeps topics and subscriptions in-sync
  • connects to all Zookeeper clusters and keeps metadata in-sync

Configuring Frontend and Consumers is easy: use configuration options from previous chapter to connect to given clusters. Remember about specifying proper kafka.cluster.name.

Since Management instances need to know all clusters, their configuration is bit more complex. Example configuration for the schematics provided above:

kafka:
  clusters:
    -
      datacenter: dc1
      clusterName: kafka_primary
      connectionString: kafka-zookeeper:2181/clusters/dc1
    -
      datacenter: dc2
      clusterName: kafka_secondary
      connectionString: kafka-zookeeper:2181/clusters/dc2

storage:
  pathPrefix: /run/hermes
  clusters:
    -
      datacenter: dc1
      clusterName: zk1
      connectionString: metadata-zookeeper.dc1:2181
    -
      datacenter: dc2
      clusterName: zk2
      connectionString: metadata-zookeeper.dc2:2181

Multiple Hermes on single Kafka cluster

It’s also possible to run multiple Hermes clusters on a single Kafka cluster, e.g. to separate different test environments. To do this, on each Hermes cluster you have to provide different value for: * kafka.namespace property in Frontend and Consumers. In Management it’s named kafka.defaultNamespace and also need to be changed. * zookeeper.root property in Frontend and Consumers if you use the same Zookeeper cluster for all Hermes clusters. In Management it’s named storage.pathPrefix and also need to be changed.

kafka.namespace property also can used to distinguish Hermes-managed topics on multi-purpose Kafka cluster.