Stopping and Starting Stand-Alone AtScale

This document describes how to stop and start Stand-Alone AtScale and AtScale services.

Before you begin

  • These instructions are for CentOS/RedHat operating systems. You may have to modify for your supported OS.
  • Unless otherwise specified, run commands as the atscale user.

About this task

You can list AtScale services, review their logs, and stop and start individual AtScale services as well as the AtScale application.

AtScale bundles Supervisor http://supervisord.org to use in managing all applications and services used to run the product. You work with AtScale's Supervisor via the atscale_service_control script.

Procedure

  • List AtScale services

    To list all the services that AtScale is managing, including their status:

    1. Run atscale_service_control status
    $ /opt/atscale/bin/atscale_service_control status
    account RUNNING pid 6011, uptime 0:01:09
    coordinator RUNNING pid 5270, uptime 0:01:47
    engine RUNNING pid 7573, uptime 0:00:24
    haproxy RUNNING pid 5371, uptime 0:01:43
    modeler RUNNING pid 5907, uptime 0:01:15
    patroni RUNNING pid 5546, uptime 0:01:35
    servicecontrol RUNNING pid 6187, uptime 0:00:57
    sidecarserver RUNNING pid 6116, uptime 0:01:01
  • Review logs

    Review the AtScale service logs.

    The logs for each of the AtScale-managed services are in the/opt/atscale/log/ directory with the naming convention <SERVICE_NAME>.stdout. Individual applications and services may have additional logging, which can be in subdirectories of /opt/atscale/log/.

  • Stop individual services

    To stop a single AtScale-managed service:

    1. Determine the name of the service via the atscale_service_control status command.

    2. Issue the stop command.

      $ /opt/atscale/bin/atscale_service_control stop engine engine: stopped

You can run this command repeatedly without impacting the services that are already stopped.

  • Stop AtScale "apps"

    To stop all the AtScale apps, which are AtScale-developed services, run the atscale_stop_apps script.

    $ /opt/atscale/bin/atscale_stop_apps
    Running this will stop all AtScale apps on this machine.
    Supporting services, such as AtScale's Postgres, will remain running.
    Continue? (y/n): y
    Stopping AtScale applications...
    account: stopped engine: ERROR (not running)
    modeler: stopped
    servicecontrol: stopped
    sidecarserver: stopped
    AtScale applications have been stopped.

You can run this command repeatedly without impacting the services that are already stopped.

  • Fully Stop AtScale

    To stop AtScale entirely, shutting down all apps and services as well as Supervisor, run theatscale_stop script.

    $ /opt/atscale/bin/atscale_stop
    Running this will fully stop all AtScale apps and services on this machine.
    Continue? (y/n): y
    Stopping all AtScale apps and services...
    Stopping AtScale applications...
    account: ERROR (not running)
    engine: ERROR (not running)
    modeler: ERROR (not running)
    servicecontrol: ERROR (not running)
    sidecarserver: ERROR (not running)
    AtScale applications have been stopped.
    patroni: stopped
    haproxy: stopped
    coordinator: stopped
    Shut down
    AtScale has been stopped.

You can run this command repeatedly without impacting the services that are already stopped.

  • Start AtScale from a full stop

    To start AtScale from a fully stopped state:

    1. As a precautionary measure before starting up AtScale, use the ps command to check if any AtScale processes still running.

      $ ps -aef | grep /opt/atscale
      atscaler 3848 3789 0 20:00 ? 00:00:07 python3
      /opt/atscale/versions/7.4.0.1233/pkg/python3/bin/patroni
      /opt/atscale/versions/7.4.0.1233/conf/database/patroni.yml
    2. Kill them if needed prior to restarting.

      $ kill -9 2149
      $ ps -aef | grep /opt/atscale
      atscaler 597 517 0 02:50 pts/0 00:00:00 grep /opt/atscale
    3. Run atscale_service_control status to determine that AtScale is completely stopped.

    4. Run the atscale_start script to start AtScale.

    $ /opt/atscale/bin/atscale_start

You can run this command repeatedly without impacting the services that are already stopped.

  1. Check the status after starting up:

    $ /opt/atscale/bin/atscale_service_control status
    account RUNNING pid 13009, uptime 0:00:11
    coordinator RUNNING pid 13024, uptime 0:00:11
    engine RUNNING pid 13013, uptime 0:00:11
    haproxy RUNNING pid 13019, uptime 0:00:11
    modeler RUNNING pid 13007, uptime 0:00:11
    patroni RUNNING pid 13003, uptime 0:00:11
    servicecontrol RUNNING pid 13015, uptime 0:00:11
    sidecarserver RUNNING pid 13014, uptime 0:00:11
  • Start AtScale apps

    To start all the AtScale apps (the AtScale-developed services), run the atscale_start_apps script.

    $ /opt/atscale/bin/atscale_start_apps
    Starting AtScale applications...
    account: started
    engine: started
    modeler: started
    servicecontrol: started
    sidecarserver: started
    AtScale applications have been started.

You can run this command repeatedly without impacting the services that are already started.

  • Start individual services

    To start a single AtScale-managed service:

    1. Determine the name of the service via the atscale_service_control status command.

    2. Run the start command.

      $ /opt/atscale/bin/atscale_service_control start engine
      engine: started

    These commands can run repeatedly without impacting the services that are already started.