Document valid TZ is required for cron

[Quietscher]

Describe the Bug

CronError: ERROR: You specified an invalid date.
    at CronTime.getNextDateFrom (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/time.js:130:19)
    at CronTime.sendAt (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/time.js:86:35)
    at CronTime.getTimeout (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/time.js:99:21)
    at CronJob.start (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/job.js:163:37)
    at new CronJob (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/job.js:68:18)
    at CronJob.from (/directus/node_modules/.pnpm/cron@4.3.4/node_modules/cron/dist/job.js:81:20)
    at scheduleSynchronizedJob (file:///directus/node_modules/.pnpm/@directus+api@file+api_@types+node@24.9.1_jiti@2.6.1_typescript@5.9.3/node_modules/@directus/api/dist/utils/schedule.js:8:25)
    at schedule (file:///directus/node_modules/.pnpm/@directus+api@file+api_@types+node@24.9.1_jiti@2.6.1_typescript@5.9.3/node_modules/@directus/api/dist/schedules/telemetry.js:22:5)
    at createApp (file:///directus/node_modules/.pnpm/@directus+api@file+api_@types+node@24.9.1_jiti@2.6.1_typescript@5.9.3/node_modules/@directus/api/dist/app.js:263:11)
    at async createServer (file:///directus/node_modules/.pnpm/@directus+api@file+api_@types+node@24.9.1_jiti@2.6.1_typescript@5.9.3/node_modules/@directus/api/dist/server.js:23:38)
2025-12-02T17:10:33: PM2 log: App name:directus id:0 disconnected
2025-12-02T17:10:33: PM2 log: App [directus:0] exited with code [1] via signal [SIGINT]

To Reproduce

Ive just upgraded.
I have a very basic directus running. (with postgres)

Directus Version

v11.13.X

Hosting Strategy

Self-Hosted (Docker Image)

Database

postgres

[linear]

CMS-1546

EDU-500

[ComfortablyCoding]

That looks like it is coming from the telemetry schedule, I cannot reproduce this by enabling telemetry with our latest docker image. Any relevant settings or changes?

[Quietscher]

Which settings do you mean?

[Quietscher]

I don't have any telemetry related env for the container

[ComfortablyCoding]

Which settings do you mean?

To clarify, by settings I meant env configuration variables. Running the following compose file with the defaults does not produce this issue. The error shown implies it is from the telemetry schedule which is registered by default, as it does not reproduce by running the container I assume something else is the cause.

services:
    directus:
        image: directus/directus:11.13.4
        ports:
            - 8055:8055
        volumes:
            - ./database:/directus/database
            - ./uploads:/directus/uploads
            - ./extensions:/directus/extensions
        environment:
            SECRET: "replace-with-random-value"
            DB_CLIENT: "sqlite3"
            DB_FILENAME: "/directus/database/data.db"
            WEBSOCKETS_ENABLED: "true"

[Quietscher]

docker run
  -d
  --name='Directus'
  --net='bridge'
  --pids-limit 2048
  -e TZ="Europe/Berlin"
  -e HOST_OS="Unraid"
  -e HOST_CONTAINERNAME="Directus"
  -e 'ADMIN_EMAIL'='admin@example.com'
  -e 'ADMIN_PASSWORD'='wikddkdk'
  -e 'DB_CLIENT'='postgres'
  -e 'DB_HOST'='192.168.1.158'
  -e 'DB_PORT'='55677'
  -e 'DB_DATABASE'='postgres'
  -e 'DB_USER'='postgres'
  -e 'DB_PASSWORD'='skekek'
  -e 'KEY'='Directus'
  -e 'SECRET'='kaks'
  -e 'PUBLIC_URL'='https://skskdk.aksopqwldn.de'
  -e 'AUTH_DISABLE_DEFAULT'='true'
  -e 'AUTH_PROVIDERS'='AUTHELIA'
  -e 'AUTH_AUTHELIA_DRIVER'='slldldlspsldnf'
  -e 'AUTH_AUTHELIA_MODE'='ksskdkbdsoks'
  -e 'CORS_ENABLED'='true'
  -e 'CORS_ORIGIN'='http://localhost:3000,https://urllll.de'
  -e 'DB_FILENAME'=''
  -e 'DB_CONNECTION_STRING'=''
  -e 'TZ'=''
  -e 'CA_TS_FALLBACK_DIR'='/directus/uploads/'
  -p '8055:8055/tcp'
  -v '/mnt/user/appdata/directus/uploads/':'/directus/uploads/':'rw'
  -v '/mnt/user/appdata/directus/extensions/':'/directus/extensions/':'rw'
  -v '/mnt/user/appdata/directus/database/':'/directus/database/':'rw'
  --restart unless-stopped 'directus/directus:11.12' 
7e2e91d88e16ca51d4a3c40255f0a75a86a35547c98d7c54d109655513a86f13

[Quietscher]

And some more auth related, but I removed those

[the-other-dev]

It is because of the empty TZ environment variable. If you set it to a valid value, e.g. UTC, Europe/Berlin, etc. Directus should start as expected.

Is there a reason for setting the TZ environment variable to an empty value?

[Quietscher]

No TZ is an env from UntaidOS not Directus

[Quietscher]

Also this did not change from directus 11.12 to 11.13

[the-other-dev]

In Directus v11.13.0, node-schedule was replaced by cron (directus/directus#25874). Unlike node-schedule, cron reads the TZ environment variable, causing the different behavior.

Disclaimer: I'm just another Directus user trying to help.

[ComfortablyCoding]

Good catch @the-other-dev , that does appear to be the issue. Setting TZ='' reproduces the problem, while using a valid or no TZ value works. I would say this is a configuration issue, so I’ll move it into our docs to make sure it’s a documented warning.

P.S. You have TZ listed twice in your docker command; TZ="Europe/Berlin" alone works.