Vector

This pack contains a single system job that runs Vector across all Nomad clients. It currently supports being run by the Docker driver.

The inbuilt configuration file is configured to read from 3 sources:

• Docker logs: (reference) plugs into the Docker socket to collect container logs;
• Host metrics: (reference) collects host metrics, such as CPU, memory, disk, and network utilization with Vector's inbuilt collector;
• Nomad metrics: (reference) collects exposed Nomad metrics via the Prometheus client.

Additionally, Docker logs will be sent to the specified Loki endpoint (with support for basic authentication) and host/Nomad metrics to the specified Prometheus endpoint (also with support for basic authentication).

Requirements

Loki and Prometheus

Running instances of Loki and Prometheus are required and are not deployed with this pack.

Vector will collect:

• Docker logs: forwarding them to a running Loki instance;
• Host metrics: forwarding them to a running Prometheus instance.

Nomad packs for Loki and Prometheus are readily available for a self-hosted solution.

Grafana Cloud, Grafana Labs' managed service, is also supported by this pack with basic authentication.

Nomad

Clients that expect to run this job require:

• Docker volumes to be enabled within their Docker plugin stanza, due to read-only bind mounts of /proc, /sys and /var/run/docker.sock:

plugin "docker" {
  config {
    volumes {
      enabled = true
    }
  }
}

• Extra Docker labels to be added within their Docker plugin stanza, in order to facilitate logs filtering:

plugin "docker" {
  config {
    extra_labels = ["job_name", "job_id", "task_group_name", "task_name", "namespace", "node_name", "node_id"]
  }
}

• Telemetry to be enabled within their config stanza:

telemetry {
  collection_interval = "5s"
  disable_hostname = false
  prometheus_metrics = true
  publish_allocation_metrics = true
  publish_node_metrics = true
}

• CNI plugins installed and its path set in the client's configuration if network mode is set to bridge.

Variables

• job_name (string "") - The name to use as the job name which overrides using the pack name.
• datacenters (list(string) ["dc1"]) - A list of datacenters in the region which are eligible for task placement.
• region (string "global") - The region where the job should be placed.
• namespace (string "default") - The namespace where the job should be placed.
• constraints (list(object)) - Constraints to apply to the entire job.
• vector_group_network (object) - The Vector network configuration options.
• vector_group_update (object) - The Vector update configuration options.
• vector_group_ephemeral_disk (object) - The Vector ephemeral disk configuration options.
• vector_task (object) - Details configuration options for the Vector task.
• vector_task_bind_mounts (object) - Details configuration options for the Vector bind mounts.
• vector_task_loki_prometheus (object) - Details Loki and Prometheus endpoints/credentials for the Vector task.
• vector_task_data_config_toml (string) - The Vector configuration to pass to the task.
• vector_task_resources (object) - The resource to assign to the Vector task.
• vector_task_services (object) - Configuration options of the Vector service and checks.

constraints List of Objects

Nomad job constraint stanza allows restricting the set of eligible nodes on which the Vector task will run.

• attribute (string) - Specifies the name or reference of the attribute to examine for the constraint.
• operator (string) - Specifies the comparison operator. The ordering is compared lexically.
• value (string) - Specifies the value to compare the attribute against using the specified operation.

The default value constrains the job to run on client whose kernel name is linux. The HCL variable list of objects is shown below and uses a double dollar sign for escaping:

[
  {
    attribute = "$${attr.kernel.name}",
    value     = "linux",
    operator  = "",
  }
]

vector_group_network Object

Nomad job network stanza specifies the networking requirements for the task group, including the network mode and port allocations.

• mode (string "bridge") - Mode of the network.
• hostname (string "${attr.unique.hostname}") - Specifies the hostname assigned to the network namespace. If hostname is not specified, the container hostname will be used in the metrics.
• ports (map<string|number> api:8686) - Specifies the port mapping for the Vector task. The map key indicates the port label, and the value is the Vector port inside the network namespace.

vector_group_update Object

Nomad job update stanza specifies the networking requirements for the task group, including the network mode and port allocations.

• min_healthy_time (string "10s") - The minimum time the allocation must be in the healthy state before it is marked as healthy.
• healthy_deadline (string "5m") - The deadline in which the allocation must be marked as healthy after which the allocation is automatically transitioned to unhealthy.
• progress_deadline (string "10m") - The deadline in which an allocation must be marked as healthy.
• auto_revert (bool "true") - Specifies if the job should auto-revert to the last stable job on deployment failure.

vector_group_ephemeral_disk Object

Nomad job ephemeral disk stanza describes the ephemeral disk requirements of the group.

• migrate (bool true) - Specifies that the Nomad client should make a best-effort attempt to migrate the data from a remote machine if placement cannot be made on the original node.
• size (number 300) - Specifies the size of the ephemeral disk in MB.
• sticky (bool true) - Specifies that Nomad should make a best-effort attempt to place the updated allocation on the same machine.

vector_task Object

• driver (string "docker") - The Nomad task driver to use to run the Vector task. Currently, only "docker" is supported.
• version (string "0.17.3-alpine") - The Vector version to run.

vector_task_bind_mounts Object

Details configuration options for the Vector bind mounts. Three directories are mounted into the container and there are 2 variables exposed for each bind mount: the source and target paths.

• source_procfs_root_path (string "/proc") - The absolute path for the procfs in the host (node).
• source_sysfs_root_path (string "/sys") - The absolute path for the sysfs in the host (node).
• source_docker_socket_path (string "/var/run/docker.sock") - The absolute path for the Docker socket in the host (node).
• target_procfs_root_path (string "/host/proc") - The absolute path for the procfs mount in the container's filesystem.
• target_sysfs_root_path (string "/host/sys") - The absolute path for the sysfs mount in the container's filesystem.
• target_docker_socket_path (string "/host/var/run/docker.sock") - The absolute path for the Docker socket mount in the container's filesystem.

vector_task_loki_prometheus Object

Details endpoints and credentials for the Loki and Prometheus instances that Vector will send data to.

• loki_endpoint_url (string "http://127.0.0.1:3100") - Loki's endpoint URL. In case the logs are to be sent to Grafana Cloud, use the base URL. For example: https://logs-prod-eu-west-0.grafana.net/.
• loki_username (string "") - Loki's basic authentication username.
• loki_password (string "") - Loki's basic authentication password.
• prometheus_endpoint_url (string "http://127.0.0.1:9090") - Prometheus's remote write endpoint URL. In case the logs are to be sent to Grafana Cloud, use the remote write URL. For example: https://prometheus-prod-01-eu-west-0.grafana.net/api/prom/push.
• prometheus_username (string "") - Prometheus' basic authentication username.
• prometheus_password (string "") - Prometheus' basic authentication password.

vector_task_resources Object

-cpu (number 64) - Specifies the CPU required to run this task in MHz. -memory (number 64) - Specifies the memory required in MB.

vector_task_services List of Objects

• service_port_label (string "api") - Specifies the port to advertise for this service.
• service_name (string "vector") - Specifies the name this service will be advertised as in Consul.
• service_tags (list(string) ["observability"]) - Specifies the list of tags to associate with this service.
• check_enabled (bool true) - Whether to enable a check for this service.
• check_path (string "/health") - The HTTP path to query Vector's API.
• check_interval (string "3s") - Specifies the frequency of the health checks that Consul will perform.
• check_timeout (string "1s") - Specifies how long Consul will wait for a health check query to succeed.

The default value for this variable configures a service for the Vector API along with a check running against the Vector API health check endpoint.

[
  {
    service_port_label = "api",
    service_name       = "vector",
    service_tags       = ["observability"],
    check_enabled      = true,
    check_path         = "/health",
    check_interval     = "3s",
    check_timeout      = "1s",
  }
]