Slowdowns over time

Not every slowdown comes from a rollout. A query slows down as a table grows, or a dependency gets slower over a few days. Hud compares every endpoint, queue and custom flow with its own normal over days and weeks, and opens an issue when one drifts well past it.

The rules

Rule	Compares	Checks	Applies to
1 Week Degradation	The past week with the week before	Every hour	Endpoints, queues and custom flows
1 Day Degradation	The past day with the week before it	Every hour	Endpoints, queues and custom flows
Latency Spike	The last 10 minutes with the day before	Every 10 minutes	Endpoints

Each rule compares P90 duration. Functions in the same flow that slowed down the same way are listed on the issue, so you can see where the slowdown started.

Defaults

By default, Hud opens an issue when an endpoint, queue or custom flow meets all three settings for its rule:

Rule	Slower by at least	Calls	P90 duration	Next issue for the same endpoint
1 Week Degradation	70%	At least 350	At least 150 ms	After 7 days
1 Day Degradation	100%	At least 100	At least 150 ms	After 7 days
Latency Spike	70%	At least 50	At least 150 ms	After 7 days

Latency Spike starts turned off, so turn it on from the Rules page if you want it. While an issue keeps recurring, Hud alerts on it again every 3 days by default. To change any of these for a service or an endpoint, open the rule on the Rules page. Detection conditions decide when Hud opens an issue, and alert conditions decide when it alerts you.

Slower than a target you set

To flag an endpoint whenever it goes over a fixed duration, set an SLO instead. See SLO rules.