Portfolio Aperture Start recording

Record and replay for backend services

Reproduce a production bug on your laptop in under a minute

Aperture records every call, query, and job your service makes, then replays one of them against the code you are editing.

Free for 14 days on up to 5,000 replays, no card. Nothing is charged unless you pick a plan.

A recorded request, replayed one step at a time

tr_8f21c4 POST /checkout/confirm checkout-api v2026.31.4 · 53 spans · 3 services · 1.2 MB
500
Fails here, step 6 of 6 gateway times out
+30.0 s after the request arrived

Recorded steps

payments.rb:87Step 6 of 6, span 53 of 53

83class Payments
84 def self.charge(total, tenders, split:)
85 body = { amount: total, tenders: tenders,
86 split: split }
87 Gateway.post("/charges", body, timeout: 30)
88 rescue Gateway::Timeout
89 raise Payments::Timeout
90 end
91end

The gateway is still waiting for the missing $76.00 when the 30 second limit passes.

Values at this step

Gateway.post
timeout 30.0 s
gateway said
awaiting $76.00
retries
0
raised
Payments::Timeout
status
500

+30,014 ms 14:02:41.352Z

Fyre Logistics cut the median time to reproduce a customer-reported bug from three days to 41 minutes, over 214 incidents. Priya Raman, staff engineer, Fyre Logistics

1.4 %median CPU cost while recording

42 msto the first span on replay

A line number is not a cause

Your error tracker names the line that threw. It does not name the row that sat in the table, the header the caller sent, or the job that ran ninety seconds earlier and left the cache in the shape that broke it. Those four facts are the bug. The line is where it surfaced.

11 calls behind the median failing endpoint at Fyre, across four services

  1. job:repriceraises the saved quote by $76.00
  2. POST /checkout/confirmreads that quote
  3. payments.rb:87times out. The tracker reports this line, and only this one
The request from the replay at the top of the page, on one axis drawn to scale. Two of the three events never reach the error report.

The recording is the reproduction

A trace is a file. Attach it to the ticket. Whoever picks the ticket up starts where you stopped, with the same rows, the same clock, and the same flag values you had.

tr_8f21c4.aperture, as it arrives in the ticket
size
1.2 MB, 53 spans, 3 services
captured
14:02:11.338Z, retained 30 days
service
checkout-api v2026.31.4 at commit 9d3ac1e
entry
POST /checkout/confirm, status 500, 30.0 s
tenant
meridian-freight-intl
outbound
payments-gw, 1 call, no answer
A redacted value, as the trace keeps it
card_number
[redacted, 16 characters]
cvv
[redacted, 3 characters]
authorization
[redacted, 71 characters]
tenders[0]
card, $300.00
tenders[1]
bank transfer, $112.00
Values matching a redaction rule never leave the machine that recorded them; the trace carries the field name and a length.

1.2 MB median trace, compressed. Bodies over 256 KB are stored by hash and fetched on replay

0 lines of instrumentation. Aperture attaches to the runtime, not to your code

Replay moves one thing at a time

Edit the code, replay the same trace, and the only variable that moved is the one you moved. Outbound calls answer from the recording, so a third party being down on Tuesday does not change Tuesday's answer.

checkout-api, on a laptop
$ aperture replay tr_8f21c4 --watch
→ clock pinned 14:02:11.338Z, 3 flag values, 53 recorded spans

checkout.rb
 59    quote = Pricing.quote_for(cart)
+60    raise TenderMismatch if tenders.sum != quote.total 61    Payments.charge(quote.total, tenders, split:)

moved   1 line, checkout.rb:60
held    every other recorded value
result  POST /checkout/confirm  500 → 409   30.0 s → 12 ms
skipped payments-gw, never called

What one recording holds

  • Inbound calls, headers, and bodies
  • Outbound calls and their answers
  • Database calls and bound values
  • Queue jobs and their payloads
  • Cache reads, writes, and misses
  • Feature-flag values as they read
  • Clock reads, to the millisecond
  • Random seeds and UUID draws
  • File reads under the app root
  • Environment values, names only
  • Thread and fiber scheduling order
  • Memory high-water mark per span

12kinds of input held for every request

256 KBthe size past which a body is stored by hash

30 daysretention, then the trace is gone

6runtimes supported, Ruby through .NET

Pay for what you replay

Every plan records without a meter. You pay for a replay once, whoever starts it.

Seats renew on the day you start and you can cancel any one of them at any time. Cancelling stops the next charge and leaves the seat working to the end of the month you have already paid for. Self-hosted runs for a year and renews the same way.

Drag to see what each plan costs at your volume. Overage lands on the next invoice. Self-hosted is $4,800 a year for the whole company, so it sits outside a per-seat line.

Solo

$19a seat, a month

Replays included
5,000
Overage a replay
$0.010
Retention
7 days
Seats
1
At 40,000 a month
$369

Renews each month. Cancel any seat at any time and the next charge stops.

Team

Recommended

$79a seat, a month

Replays included
50,000
Overage a replay
$0.006
Retention
30 days
Seats
2 to 40
At 40,000 a month
$79

Renews each month. Cancel any seat at any time and the next charge stops.

Example team
Meridian Freight Systems International
Seats
12 of up to 40
At this volume
12 × $79 = $948

Self-hosted

$4,800a year, unmetered

Replays included
Unmetered
Overage a replay
None
Retention
Yours
Seats
Unlimited
Averaged over a month
$400

Renews each year. Cancel at any time and the next charge stops.

Questions before you start

What counts as a replay?

One run of a recording against your code, by anyone on the team. Reading a trace in the debugger does not count. A replay that ends in an error still counts, because the work was done.

What does recording cost my service?

A median of 1.4 % of CPU while recording. On replay, the first span appears in 42 ms.

Does card data leave our machines?

No. A value matching a redaction rule stays on the machine that recorded it. The trace keeps the field name and a length, such as card_number, 16 characters.

What happens when the trial ends?

After 14 days or 5,000 replays, whichever comes first, replay pauses until you pick a plan. No card was taken, so nothing is charged.

Can we run it inside our own network?

Self-hosted is $4,800 a year for the whole company: unmetered, unlimited seats, and the retention is yours to set.

The first replay is the one that decides it

Install the agent on one service and record a single endpoint. If the first bug you replay already has a fix, you lose an afternoon.

Terminal
$ brew install aperture
$ aperture attach checkout-api
→ recording checkout-api, 0 lines changed

Trial
14 days
Replays
5,000
Card
Not asked for
Services
One is enough

How this page was decidedDesign notes

Shape of the page
An editorial spread, with the product itself as the first thing on it. The argument hangs its numeral in one margin and its evidence in the other.
What it leads with
A recorded request you can scrub. Every other figure sits inside a sentence, or in a margin line that names what it counts.
Spacing
Generous, and broken deliberately in one place: the recording list is set close, so that the space between movements does the separating instead.
The one bold decision
Every colour on the page was measured before it was used. None of Tailwind's popular defaults survived: not the stock indigo, not the pale grey body text, not the pale grey control border.
Where the colour comes from
Tailwind's own published palette, with each step checked against the surface it lands on. The code window is dark in both themes, so its syntax colours were measured once, against that one surface.
Type
The three families Tailwind ships and nothing bought in: a sans for the interface, a serif for the movement numerals, and a typewriter face for every figure and every line of code, set so columns of numbers line up.
7.56:1Grey on white, every muted line
13.71:1Light teal on near-black, the dark accent
Readability, measured. Higher is easier to read, and 4.5 is the accessibility floor.
What sits on whatCarriesRatio
White on deep tealthe main button7.55:1
Grey on whiteevery muted line7.56:1
Teal on whitelinks, numerals, focus ring5.44:1
Pale grey on near-blackthe code window16.28:1
Light teal on near-blackthe dark accent13.71:1
Near-black on light tealthe dark main button13.71:1
Grey on near-blackthe dark muted line7.68:1
Grey on the code windowcomments, line numbers6.79:1
Pale grey on dark redthe failing line14.72:1
Red on whitethe over-volume state4.92:1

Two rules, not one. The faint line that separates rows is far too pale to read text against, so it separates rows and does nothing else. Muted text is a step darker than the accessibility standard strictly asks for, because a second, newer standard reads that lighter step as usable only at large sizes. And of Tailwind's twenty-six hues, only six are readable on white at the middle of their scale. Teal is not one of them, so this page uses it two steps deeper.

Start your 14-day trial

Free for 14 days on up to 5,000 replays. No card is taken, so nothing is charged.

After the trial
Renews
To stop