Requizon vs Laravel Pulse

Laravel Pulse is Laravel's free, self-hosted performance dashboard. One of its cards is Slow Outgoing Requests, which lists the HTTP client calls that took longer than a threshold. If you have Pulse installed, you already have a view of your outbound traffic, and it is fair to ask what Requizon would add. The short version is failures: Pulse measures how slow a call was, and never records whether it worked.

What the Slow Outgoing Requests recorder keeps#

Pulse's SlowOutgoingRequests recorder registers global middleware on Laravel's HTTP client. When a call finishes, successfully or not, it compares the duration with a threshold: 1,000 ms by default, adjustable per URL pattern. Calls under the threshold are discarded. For the rest it records a count and the slowest duration, keyed by the HTTP method and the URL, after applying any grouping patterns you have configured.

That is the whole record. The card shows each method and URL with how many times it was slow and its slowest time, over the last hour, six hours, 24 hours or seven days. Pulse trims data older than seven days by default.

At a glance#

The calls Pulse cannot show you#

A threshold on duration is a good filter for slowness and the wrong filter for failure, because plenty of failing calls are fast. An expired API key can get its 401 in 60 ms. A vendor's broken deploy returns 500s as fast as it returns anything. A misconfigured DNS name fails before a connection is even opened. None of those reach the threshold, so none of them appear on the card, and an API that is failing every call can look perfectly healthy in Pulse.

The calls that do reach the card are recorded without a status, so a slow success and a slow timeout land in the same row, and telling them apart means going to find the exception.

Requizon records every transfer with its outcome. The overview charts responses per API stacked by status code or failure type, so the hour a vendor started returning 503s shows up as a change of colour in the chart, whatever the latency did.

Where Pulse is ahead#

• It costs nothing, it is maintained by the Laravel team, and there is a good chance it is already in your composer.json.
• One screen for the whole app. Slow queries, slow jobs, exceptions and server load next to slow outgoing requests is a better morning check than any single-purpose dashboard.
• It runs on PostgreSQL, and on Laravel 10 and 11. Requizon needs MySQL and Laravel 12 or later.
• It scales down its own cost. Sampling and a Redis ingest are built in for high-traffic applications. Requizon's answer to volume is its rollup and pruning, and it writes one row per transfer before that.
• Slowness is sometimes the only question. If your integrations fail loudly in your error tracker already and you just want to know which one drags, the Slow Outgoing Requests card answers that.

Grouping URLs#

Pulse groups by the full URL unless you give it regular expressions, so /orders/1 and /orders/2 are separate rows until you write a pattern such as '#/\d+#' => '/*'. It is flexible, and the patterns are yours to maintain as APIs change.

Requizon normalises paths without configuration: numeric, UUID, ULID and hex segments, and any segment over 40 characters, become :id. Hosts can be mapped to API names, so several hostnames for one vendor share a row. When the heuristic is wrong for an API you can set allowed patterns or replace the normaliser.

Keeping Pulse and adding Requizon#

Both register global middleware on Laravel's HTTP client factory and neither changes the request, so they run together without any configuration. Keep Pulse as the overview of the application, and look at Requizon when the question is whether a particular API is failing and since when.