This project contains Docker image build instructions. This image is meant to facilitate the deployment of Apache ZooKeeper on Kubernetes using StatefulSets. You may also want to check the corresponding helm chart.
To give it a quick try, you can either build image locally:
$ make build
$ make run
or pull and run existing image from the DockerHub:
$ docker run -it --rm --net=host svlady/alpine-zookeeper
- Scaling up and down the ensemble is not supported in this Zookeeper version. Operators may resort to "rolling restarts" - a manual and error-prone method of changing the configuration that could cause data loss and inconsistency in production. Alternatively, you may want to take a look at 3.5.x branch, which is supporting dynamic configuration.
- The Observer feature is currently not supported by this project.
- Netty and SSL/TLS settings are currently not supported by this project.
The docker image that may be built out of this repository is based on Alpine linux and includes the latest release of the OpenJDK JRE based on the 1.8 JVM, as well as the latest stable release of ZooKeeper.
The image is providing a custom entrypoint which is dropping privileges and launches
Zookeeper process as a non-root user (zookeeper). The Zookeeper package is installed
into the /opt/zookeeper
directory by default. This installation path can be
specified as an argument during the package build.
The bootstrap.sh
script, called by the entrypoint, generates the following files
using sane defaults and environment variables that are injected into the container:
- the ZooKeeper configuration (zoo.cfg)
- Log4J configuration (log4j.properties)
- and JVM configuration (jvm.env)
These files are located in the /opt/zookeeper/conf
directory by default.
This is a mandatory configuration variable that is used to configure the membership of the Zookeeper ensemble. It is also used to prevent data loss during accidental scale operations.
Variable | Type | Default | Description |
---|---|---|---|
ZK_SERVERS | string | N/A | A colon separated list of servers in the ensemble, e.g. server.x=[hostname]:nnnnn[:nnnnn];... |
For example:
ZK_SERVERS=zk-0.zookeeper-headless:2888:3888;zk-1.zookeeper-headless:2888:3888;zk-2.zookeeper-headless:2888:3888
Please note: unlike Zookeeper connection string, this list is using a semicolon as a separator.
The ZK_CLIENT_PORT
, ZK_ELECTION_PORT
, and ZK_SERVERS_PORT
must be set
equal to the container ports specified in the container configuration, and
the ZK_SERVER_PORT
and ZK_ELECTION_PORT
must also match the Headless
Service configuration.
Variable | Type | Default | Description |
---|---|---|---|
ZK_CLIENT_PORT | integer | 2181 | The port on which the server will accept client requests. |
ZK_SERVER_PORT | integer | 2888 | The port on which the leader will send events to followers. |
ZK_ELECTION_PORT | integer | 3888 | The port on which the ensemble performs leader election. |
ZK_MAX_CLIENT_CNXNS | integer | 60 | The maximum number of concurrent client connections that a server in the ensemble will accept. |
These environment variables may be omitted and the values specified above will be used by default.
There is no need to modify these variables for the most use-cases. Please note that the minimum session timeout will be 2 ticks.
Variable | Type | Default | Description |
---|---|---|---|
ZK_TICK_TIME | integer | 2000 | The number of wall clock ms that corresponds to a Tick for the ensembles internal time. |
ZK_INIT_LIMIT | integer | 5 | The number of Ticks that an ensemble member is allowed to perform leader election. |
ZK_SYNC_LIMIT | integer | 10 | The number of Ticks by which a follower may lag behind the ensembles leader. |
These environment variables may be omitted and the values specified above will be used by default.
Variable | Type | Default | Description |
---|---|---|---|
ZK_MIN_SESSION_TIMEOUT | integer | 2 * ZK_TICK_TIME | The minimum session timeout that the ensemble will allow a client to request. |
ZK_MAX_SESSION_TIMEOUT | integer | 20 * ZK_TICK_TIME | The maximum session timeout that the ensemble will allow a client to request. |
These environment variables may be omitted and the values specified above will be used by default.
If you do not have an existing retention policy and backup procedure, and if you are comfortable with an automatic procedure, you can use the environment variables below to enable and configure automatic data purge policies.
Variable | Type | Default | Description |
---|---|---|---|
ZK_SNAP_RETAIN_COUNT | integer | 3 | The number of snapshots that the ZooKeeper process will retain if ZK_PURGE_INTERVAL is set to a value greater than 0. |
ZK_PURGE_INTERVAL | integer | 0 | The delay, in hours, between ZooKeeper log and snapshot cleanups. |
These environment variables may be omitted and the values specified above will be used by default.
Please note: Zookeeper does not, by default, purge old transactions logs or
snapshots. This can cause the disk to become full. If you have backup procedures
and retention policies that rely on external systems, the snapshots and data logs
can be retrieved manually from the /opt/zookeeper/data
and /opt/zookeeper/logs
directories correspondingly. These will be stored on the persistent volume. The
zkCleanup.sh
script can be used to manually purge outdated logs and snapshots.
One of the most important settings is JVM heap size. Make sure that the heap size you request does not cause the process to swap out or being killed due to OOM Exception, when reaching the limit alotted to container or pod.
Variable | Type | Default | Description |
---|---|---|---|
ZK_HEAP_SIZE | integer | 512M | The JVM heap size. |
ZK_JMX_PORT | integer | N/A | The JMX port used for monitoring and metric collection. |
ZK_JVM_FLAGS | string | "-server -XX: UseG1GC -XX:MaxGCPauseMillis=20 -XX:InitiatingHeapOccupancyPercent=35 -XX: ExplicitGCInvokesConcurrent -Djava.awt.headless=true -Djava.net.preferIPv4Stack=true" | Server settings for JVM. |
ZK_JVM_DEBUG | string | N/A | When variable is set to any non-empty value JVM is running in debug mode. |
ZK_JVM_DEBUG_PORT | integer | 5005 | JVM debug port. |
ZK_JVM_DEBUG_SUSPEND_FLAG | enum(n,y) | n | JVM is not suppended. |
ZK_JVM_GC_LOG | string | N/A | Set to the path of the GC log file. |
These environment variables may be omitted and the values specified above will be used by default.
The ZooKeeper process must be run in the foreground, and the log information will
be shipped to the stdout or stderr. This is considered to be a best practice for
containerized applications. The following variable ZK_LOG_LEVEL
is providing a
threshold for the log messages sent to the console.
Shall you need to fine-tune logging facilities for specific log producers, you
may use ZK_LOG4J_LOGGERS
variable to specify different log levels and
thresholds as defined by LOG4J specification.
For example, you may use the following configuration to drop log messages caused by each client connection due to pod liveness and readiness checks.
ZK_LOG4J_LOGGERS=org.apache.zookeeper.server.NIOServerCnxnFactory=WARN,org.apache.zookeeper.server.NIOServerCnxn=WARN
Variable | Type | Default | Description |
---|---|---|---|
ZK_LOG_LEVEL | enum(TRACE,DEBUG,INFO,WARN,ERROR,FATAL) | INFO | The Log Level for the Zookeeper processes logger. |
ZK_LOG4J_LOGGERS | comma separated string | N/A | The Log Level for specific logger. |
The recommmended way to check Zookeeper readiness is to use one of so called
4wl commands.
Healthy Zookeper shall answer imok
to the ruok
request sent to the
client port. The example below demonstrates how to configure liveness
and readiness probes for the Pods in the Stateful Set.
livenessProbe:
exec:
command: ['/bin/sh', '-c', 'echo "ruok" | nc -w 2 localhost 2181 | grep imok']
initialDelaySeconds: 10
timeoutSeconds: 5
readinessProbe:
exec:
command: ['/bin/sh', '-c', 'echo "ruok" | nc -w 2 localhost 2181 | grep imok']
initialDelaySeconds: 10
timeoutSeconds: 5
There is a little caveat, though. By default, Zookepeer is accepting all commands and for number of reasons you may want to allow only specific commands to be accepted. You can use the following setting to whitelist only desired commands and ignore the rest.
ZK_CMD_WHITELIST="stat, ruok"
Variable | Type | Default | Description |
---|---|---|---|
ZK_CMD_WHITELIST | comma separated string | * | Allowed 4wl commands. |
The image is assuming two volumes:
data
used to persist Zookeper database, located by default at/opt/zookeeper/data
logs
used to persist Zookeper transaction logs, located by default at/opt/zookeeper/logs
It is recommended to keep those volumes separate in order to avoid I/O contention and improving Zookeper performance.
The volume mounts for the container may be defined as shown below.
volumeMounts:
- name: dataDir
mountPath: /opt/zookeeper/data
- name: dataLogDir
mountPath: /opt/zookeeper/logs
For production ready user-cases, the use of Persistent Volumes is mandatory. Please note: If you use the image with emptyDirs, you will likely suffer a data loss.
The example below demonstrates how to request a dynamically provisioned persistent volume of 20 GiB.
volumeClaimTemplates:
- metadata:
name: datadir
annotations:
volume.alpha.kubernetes.io/storage-class: anything
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 20Gi