DataStax Enterprise recommended settings for Docker
Follow the recommended guidance and settings for using DataStax Enterprise (DSE) with Docker.
To ensure your success when using Docker, follow the recommended guidance and settings for using DataStax Enterprise (DSE) with Docker.
|Although DataStax provides the following guidance, adaptations of these instructions might be required depending on the deployment. It is highly recommended to rigorously test the use cases under consideration before deploying a DataStax installation on Docker in production environments.|
DSE achieves resilience and high availability through a cluster of nodes that replicate data across the cluster. This replication ensures that if any individual node fails, access to data is not lost and performance is maintained. However, in a containerized environment, running multiple DSE nodes on the same physical hardware will introduce a single point of failure.
|To avoid a single point of failure, run only a single DataStax container on a DSE cluster per Docker host. If running multiple DataStax containers on a single Docker host, ensure that the containers are in different DSE clusters.|
The official DataStax images include the latest DataStax Agent version at the time of official image build. If you require a version of the DataStax Agent that differs from the one included with the official image, you must build an image that includes the required versions.
For minimum container resource requirements, follow the capacity planning guidance for selecting hardware for production environments:
The default SSD configurations on most Linux distributions are not optimal. To ensure the best settings, see the recommended production settings to optimize SSDs:
readahead setting for RAID on SSDs (in Amazon EC2) is 8 KB, the same as it is for non-RAID SSDs.
For details, see Optimizing SSDs.
readahead of 128 is recommended.
Check to ensure
setra is not set to 65536:
sudo blockdev --report /dev/spinning_disk
To set setra:
sudo blockdev --setra 128 /dev/spinning_disk
Because time is not namespaced in the Linux kernel, containers share the clock with the Docker host machine. Ensure that clocks are synchronized on the host machines and containers by configuring NTP or other methods on the host machines.
Swapping must be disabled for performance and node stability. Run the following command on the Docker host to disable swap. The Docker host passes this setting to the container.
sudo swapoff --all
To disable swap per container, see Preventing a container from using SWAP in the Docker documentation.
To make this change permanent, remove all swap file entries from
To ensure optimal performance, do not use governors that lower the CPU frequency.
Instead, reconfigure all CPUs to use the
performance governor on the Docker hosts.
for CPUFREQ in /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor do [ -f $CPUFREQ ] || continue echo -n performance > $CPUFREQ done
THP can cause performance issues in DSE when it defragments 4 K chunks into 2 MB chunks.
defrag, run the following command on the Docker host:
echo never | sudo tee /sys/kernel/mm/transparent_hugepage/defrag
All containers by default inherit user limits from the Docker daemon.
In production environments, DSE expects the following changes to
ulimit -n 100000 # nofile: max number of open files ulimit -l unlimited # memlock: maximum locked-in-memory address space
Run the following command to check the Docker daemon defaults for
docker run --rm ubuntu /bin/bash -c 'ulimit -a'
ulimitfor Docker containers, run the
docker runcommand with the following
--ulimit nofile=100000:100000 --ulimit nproc=32768 --ulimit memlock=-1:-1
DSE tries to lock memory using
When running in Docker, that capability is disabled.
mlock, add the following option to the
docker run command:
On the Docker host, check the value of
vm.max_map_count, which should be set to 1048575.
To set the value of
vm.max_map_count, add the following line to /etc/sysctl.conf, and then run
sysctl -p to propagate the changes.
vm.max_map_count = 1048575
sudo sysctl -p
For each container in production environments, explicitly set the JVM heap size using the
JVM_EXTRA_OPTS environment variable with the
docker run command.
For example, to use 16 GB for the JVM heap, run the
docker run command with the following option:
docker run -e JVM_EXTRA_OPTS="-Xms16g -Xmx16g"
For advanced configuration management, DataStax provides a mechanism for modifying configurations without replacing or customizing DataStax Docker containers. When any of the approved configuration files are mounted to a host volume, the files are mapped automatically within the container. See Using the DSE configuration volume.
The DSE Docker container writes all node-specific data in the directories under
/var/lib/cassandra/ by default.
To persist this data, map the data directories inside the container to a directory on the host file system using the
-v option with the
docker run command, or by using a volume driver.
For example, to mount the DSE data volume to the
/dse/data directory on the Docker host, run the
docker run command with the following option:
docker run -v /dse/data:/var/lib/cassandra
/var/lib/cassandra directory outside the container with the
-v option allows the container to be deleted and recreated without losing data.
See Persisting data.
If using the Docker
devicemapper storage driver, do not use the default
loop-lvm mode, which is only appropriate for testing.
docker-engine to use direct-lvm mode, which is suitable for production environments.
When running Docker for Windows, the default resources allocated to the Linux VM running docker are 2 GB RAM and 2 CPUs. Adjust these resources as appropriate to meet the requirements for your containers. See Getting Started in Docker Desktop for Windows.
Because the default network settings in Docker (via Linux bridge) slows networking considerably, do not use these network settings in production environments.
Instead, use docker host networking by adding the
--network host option to the
docker run command, or use a plugin that can manage IP ranges across clusters of hosts.
The host networking limits the number of nodes per Docker host to one, which is the recommended configuration to use in production.
docker run -d --network host --name container_name