open-nomad/website/source/docs/jobspec/environment.html.md

4.9 KiB

layout page_title sidebar_current description
docs Runtime Environment docs-jobspec-environment Learn how to configure the Nomad runtime environment.

Runtime Environment

Some settings you specify in your job specification are passed to tasks when they start. Other settings are dynamically allocated when your job is scheduled. Both types of values are made available to your job through environment variables.

Summary

Variable Description
NOMAD_ALLOC_DIR Path to the shared alloc directory
NOMAD_TASK_DIR Path to the local task directory
NOMAD_MEMORY_LIMIT The task's memory limit in MB
NOMAD_CPU_LIMIT The task's CPU limit in MHz
NOMAD_ALLOC_ID The allocation ID of the task
NOMAD_ALLOC_NAME The allocation name of the task
NOMAD_ALLOC_INDEX The allocation index; useful to distinguish instances of task groups
NOMAD_TASK_NAME The task's name
NOMAD_IP_"label" The IP of the the port with the given label
NOMAD_PORT_"label" The port value with the given label
NOMAD_ADDR_"label" The IP:Port pair of the the port with the given label
NOMAD_HOST_PORT_"label" The host port for the given label if the port is port mapped
NOMAD_META_"key" The metadata of the task

Task Identifiers

Nomad will pass both the allocation ID and name as well as the task's name. These are given as NOMAD_ALLOC_ID, NOMAD_ALLOC_NAME, NOMAD_ALLOC_INDEX and NOMAD_TASK_NAME. The allocation ID and index can be useful when the task being run needs a unique identifier or to know its instance count.

Resources

When you request resources for a job, Nomad creates a resource offer. The final resources for your job are not determined until it is scheduled. Nomad will tell you which resources have been allocated after evaluation and placement.

CPU and Memory

Nomad will pass CPU and memory limits to your job as NOMAD_CPU_LIMIT and NOMAD_MEMORY_LIMIT. Your task should use these values to adapt its behavior to fit inside the resource allocation that nomad provides. For example, you can use the memory limit to inform how large your in-process cache should be, or to decide when to flush buffers to disk.

Both CPU and memory are presented as integers. The unit for CPU limit is 1024 = 1Ghz. The unit for memory is 1 = 1 megabytes.

Writing your applications to adjust to these values at runtime provides greater scheduling flexibility since you can adjust the resource allocations in your job specification without needing to change your code. You can also schedule workloads that accept dynamic resource allocations so they can scale down/up as your cluster gets more or less busy.

Networking

Nomad assigns IPs and ports to your jobs and exposes them via environment variables. See the Networking page for more details.

Task Directories

Nomad makes the following two directories available to tasks:

  • alloc/: This directory is shared across all tasks in a task group and can be used to store data that needs to be used by multiple tasks, such as a log shipper.
  • local/: This directory is private to each task. It can be used to store arbitrary data that shouldn't be shared by tasks in the task group.

Both these directories are persisted until the allocation is removed, which occurs hours after all the tasks in the task group enter terminal states. This gives time to view the data produced by tasks.

Depending on the driver and operating system being targeted, the directories are made available in various ways. For example, on docker the directories are binded to the container, while on exec on Linux the directories are mounted into the chroot. Regardless of how the directories are made available, the path to the directories can be read through the following environment variables: NOMAD_ALLOC_DIR and NOMAD_TASK_DIR.

Meta

The job specification also allows you to specify a meta block to supply arbitrary configuration to a task. This allows you to easily provide job-specific configuration even if you use the same executable unit in multiple jobs. These key-value pairs are passed through to the job as NOMAD_META_"key"={value}, where key is UPPERCASED from the job specification.

Currently there is no enforcement that the meta values be lowercase, but using multiple keys with the same uppercased representation will lead to undefined behavior.