Odd Cron format doesn't appear to be interpreted properly

[elreydetoda]

So, I've got a niche (or at least from a cron perspective) use case that I need to run a playbook on every first monday of the month.

I found this article which talks about how you can override the default behavior in cron to do this. So, I set the cron in semaphore to be: 0 5 */100,1-7 * MON

Unfortunately it seems like semaphore isn't interpreting that syntax correctly, because it just ran on this past monday instead of 2023-03-06 05:00:00.

It looks like it accepted the string properly when I dump out the schedule table in the web-ui & DB:

web-ui:

DB:

> select * from project__schedule where project_id = 1 AND template_id = 13;
+----+-------------+------------+---------------------+---------------+------------------+
| id | template_id | project_id | cron_format         | repository_id | last_commit_hash |
+----+-------------+------------+---------------------+---------------+------------------+
|  2 |          13 |          1 | 0 5 */100,1-7 * MON |          NULL | NULL             |
+----+-------------+------------+---------------------+---------------+------------------+

This was the event table:

> select * from event where id > 1030 and id < 1045;

| 1037 |          1 |       192 | task        | Task ID 192 queued for running                           | 2023-02-20 05:00:01 |    NULL |
| 1038 |          1 |       192 | task        | Task ID 192 (Monthly Updates - <REDACTED>) is preparing       | 2023-02-20 05:00:01 |    NULL |
| 1039 |          1 |       192 | task        | Task ID 192 (Monthly Updates -  <REDACTED>) finished - WAITING | 2023-02-20 05:00:09 |    NULL |
| 1040 |          1 |       192 | task        | Task ID 192 (Monthly Updates -  <REDACTED>) is running         | 2023-02-20 05:00:11 |    NULL |
| 1041 |          1 |       192 | task        | Task ID 192 (Monthly Updates -  <REDACTED>) finished - ERROR   | 2023-02-20 06:01:17 |    NULL |

I also looked at the systemd log and didn't see anything in there between those time:

# journalctl -eu semaphore --since "2023-02-20 04:00:01" --until "2023-02-20 07:01:17" --output=cat --no-pager
time="2023-02-20T05:00:00Z" level=info msg="Task 192 added to queue"
time="2023-02-20T05:00:01Z" level=info msg="Set resource locker with TaskRunner 192"
time="2023-02-20T05:00:08Z" level=info msg="Stopped preparing TaskRunner 192"
time="2023-02-20T05:00:08Z" level=info msg="Release resource locker with TaskRunner 192"
time="2023-02-20T05:00:11Z" level=info msg="Set resource locker with TaskRunner 192"
time="2023-02-20T05:00:11Z" level=info msg="Task 192 removed from queue"
time="2023-02-20T06:01:17Z" level=info msg="Stopped running TaskRunner 192"
time="2023-02-20T06:01:17Z" level=info msg="Release resource locker with TaskRunner 192"

---

Is there some other places I should look for errors or issues?

Please let me know if I can help in any other way.

While this shouldn't matter here is some more info on the setup:

• Server: AWS Ubuntu 22.04 EC2 instance
• DB: AWS RDS MySQL 5.7

• Semaphore Service: running as systemd service

  # /etc/systemd/system/semaphore.service
  [Unit]
  Description=Semaphore Ansible
  Documentation=https://github.com/ansible-semaphore/semaphore
  Wants=network-online.target
  After=network-online.target
  
  [Service]
  Type=simple
  ExecReload=/bin/kill -HUP $MAINPID
  ExecStart=/usr/local/sbin/semaphore service --config=/etc/semaphore.json
  SyslogIdentifier=semaphore
  Restart=always
  
  [Install]
  
  WantedBy=multi-user.target

• config

  {
      "mysql": {
              "host": "<REDACTED>",
              "user": "<REDACTED>",
              "pass": "<REDACTED>",
              "name": "<REDACTED>",
              "options": null
      },
      "bolt": {
              "host": "",
              "user": "",
              "pass": "",
              "name": "",
              "options": null
      },
      "postgres": {
              "host": "",
              "user": "",
              "pass": "",
              "name": "",
              "options": null
      },
      "dialect": "mysql",
      "port": "",
      "interface": "",
      "tmp_path": "/opt/ansible",
      "cookie_hash": "<REDACTED>",
      "cookie_encryption": "<REDACTED>",
      "access_key_encryption": "<REDACTED>",
      "email_sender": "",
      "email_host": "",
      "email_port": "",
      "email_username": "",
      "email_password": "",
      "web_host": "",
      "ldap_binddn": "",
      "ldap_bindpassword": "",
      "ldap_server": "",
      "ldap_searchdn": "",
      "ldap_searchfilter": "",
      "ldap_mappings": {
              "dn": "",
              "mail": "",
              "uid": "",
              "cn": ""
      },
      "telegram_chat": "",
      "telegram_token": "",
      "slack_url": "<REDACTED>",
      "max_parallel_tasks": 0,
      "email_alert": false,
      "email_secure": false,
      "telegram_alert": false,
      "slack_alert": true,
      "ldap_enable": false,
      "ldap_needtls": false,
      "ssh_config_path": "",
      "demo_mode": false
  }

[Caesarovich]

Hi, Semaphore uses the Go library robfig/cron to handle cron jobs. When setting up cron schedules you have a little tooltip linking you to this package's supported cron formats (see here).

So it appears that you are trying to use a cron format unsupported by this library.

[EHEX-schildt]

You may want to test a newer Version Cron and Schedules got improved.