Docs > Platform Observability > Getting Started with Storage & Database

Getting Started with Storage & Database

Overview

This guide introduces AppStatus Storage & Data Layer and covers:

Disk capacity and growth rate per volume
Database statements ranked by total time, not single-run time
Cache hit ratio as an early warning for latency
Message queue depth, lag and drain estimate
Alerting on trend, not just on a single threshold

What is Storage & Data Layer in AppStatus?

Storage covers the stateful half of your stack: the disks that fill up, the queries that get slower as data grows, and the queues that back up when a consumer falls behind.

Database statements are normalised before ranking, so one slow query shape appears as a single row with its total cost, rather than thousands of near-identical entries you have to group by eye.

Key capabilities:

Volume capacity with growth rate, not just percentage full
Query performance ranked by total time across all executions
Cache hit ratio trend per database
Queue depth and consumer lag with a drain estimate
A trace-derived query view that needs nothing installed on the database

What you will see

What the slow-query view looks like

app.appstatus.ioSample data

Statement	Calls	Total time	Mean	Share
SELECT … FROM orders WHERE user_id = ?	1.2M	48 min	2.4 ms	31%
UPDATE sessions SET seen_at = ?	4.8M	31 min	0.4 ms	20%
SELECT … FROM products JOIN …	18k	22 min	73 ms	14%
DELETE FROM audit_log WHERE …	96	11 min	6.9 s	7%

Scroll the table sideways to see every column.

The top row is fast per call and still your biggest cost. Rank by total time, not mean.

Troubleshooting

A disk alert fires at 80% every week and nothing happens

Alert on growth rate as well as absolute usage. A volume steady at 80% is fine; one that climbed from 40% in three days needs attention even though it is at the same number.

Queue depth is high but nothing seems broken

Check lag and the drain estimate. A queue that is deep but draining steadily is healthy — one that is shallow and growing is not. Depth alone does not tell you which you have.

Query latency rose with no code change

Look at cache hit ratio first. A falling ratio as data grows produces exactly this — slower queries with identical SQL — and it predicts the problem before users feel it.

Operational Guidance

Query statements are normalised before ranking, so one slow shape shows as one row, not thousands.
A falling cache hit ratio predicts a latency problem before users feel it.
A queue that is deep but draining is healthy; one that is shallow and growing is not.

Step-by-Step Setup

Storage covers three different things that fail differently — disks, queries and queues. Set each one up against trend rather than a single number, because all three warn you well before they break if you are measuring the right thing.

Before you start

The agent installed for disk metrics
A connected database or queue for the other views

1
Confirm volumes report
Open the storage view and confirm each host lists its volumes with usage and growth.
WhereStorage → Volumes
2
Alert on growth, not just usage
Set capacity alerts on rate of change alongside absolute percentage. A volume steady at 80% is fine; one that climbed from 40% in three days is not, and a percentage-only alert cannot tell them apart.
WhereStorage → Thresholds
Tip
This is the single change that stops disk alerts from being ignored.
3
Review the slow-query list
Open the query view and read it ranked by total time. Statements are normalised first, so one slow shape appears once with its true total cost rather than as thousands of near-identical rows.
WhereStorage → Queries
4
Add queue thresholds
For anything with a consumer, alert on sustained depth with a rising trend rather than on a single spike. Compare against lag — deep but draining is healthy, shallow but growing is not.
WhereStorage → Queues

Configuration Options

Every option you can set, what each choice means, and what to pick. Use this as a reference while you fill in the form.

What to alert on

Field	Options	What it does	Recommended
Volume usage	Percentage	How full a disk is.	Pair with growth rate or it will be ignored.
Growth rate	Per day	How fast it is filling.	The number that actually predicts an outage.
Query ranking	Total / mean time	How slow statements are ordered.	Total time. A fast query run millions of times is your real cost.
Queue lag	Messages / time	How far consumers trail producers.	Lag matters more than depth for streaming systems.

Feature Reference

Every feature, where to find it in the app, and what it does. Use this when you know what you want to do but not where it lives.

Feature	Where in app	Description
Volume capacity	Storage → Volumes	Usage and growth rate per mount.
Query performance	Storage → Queries	Normalised statements ranked by total time, with cache hit ratio.
Queue health	Storage → Queues	Depth, lag and an estimate of how long a backlog takes to drain.
Trace-derived queries	APM → trace	Query timing from traces, needing nothing installed on the database.

Next Steps

Continue building your monitoring stack:

APM & Tracing

See which request runs the slow query.

Infrastructure

Capacity on the hosts underneath.

Metrics

Chart any of this over a longer window.