actinity/laravel-queue-status

There is no license information available for the latest version (3.3) of this package.

Queue Status Monitor

3.3 2024-10-22 15:45 UTC

This package is auto-updated.

Last update: 2024-11-13 10:06:27 UTC


README

Once installed this will add itself to the schedule and dispatch a job every minute on the default queue which updates a cached value.

The /queue-status-monitor endpoint shows the status of the monitor and will return either 200 OK or 400 if queues have not run within the expected threshold (5 minutes by default. This endpoint can then be monitored by Pingdom or similar.

If you are using multiple queues, or if you wish to customise thresholds you can add a monitor array to config/queue.php. Examples are:

Monitor a single queue

    "monitor" => "myqueue"

Monitor multiple queues

    "monitor" => [
        "myqueue",
        "myotherqueue"
    ]

Monitor a queue with a custom threshold

    "monitor" => [
        ["name" => myqueue","threshold" => 600],
    ]

Threshold values are specified in seconds and represent the point where the monitor will report that the queue is failing if the job has not run within that period.

Delay

As well as reporting the last run time, the /queue-status-monitor endpoint also reports the delay. This is the time between the job being dispatched and handled. This does not impact the overall status and will not trigger failures. It is purely for reference.

Failed Jobs

By default, if you have a failed_jobs table, then any failures will be treated as errors and generate a 400 status code. You can disable this by adding the following to config/queue.php.

"do_not_monitor_failed_jobs" => true

Alternatively, if you are prepared to tolerate failed jobs, but don't want them to stop you knowing if there is another issue; you can monitor independent endpoints rather than just the default:

queue-status-monitor/without-failed will ignore failed jobs and only report on thresholds and misconfiguration.

queue-status-monitor/only-failed will only check failed jobs.

Timeout vs retry_after misconfiguration checks

Queue workers are typically configured via Supervisor or similar, and the queue:work command includes a --timeout property. If this is set higher than the corresponding retry_after property (in config.queue) it can cause jobs to be duplicated. The monitor checks for this and will alert in the endpoint (and throw a 400 response).

Last dispatch

The Last dispatch value at the top of the queue monitor indicates the last time the queue-status:ping command was dispatched by the scheduler (typically every minute if you're using standard Laravel cron configuration). This doesn't trigger any errors by itself, but is shown for reference to help debug if the issue is with your crontab.

Auth

If you want to protect your monitor endpoints, add:

"status_password" => "YOURPASSWORD"

To the queue config. The default username is 'queues', but you can customise this as well by setting:

"status_user" => "USERNAME"

Disable in certain environments

You can disable the queue monitor by adding

"status_monitor_disabled" => env("QUEUE_MONITOR_DISABLED",false)

in your queue config, and then setting that value to true in your .env file.