What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
Cloudflare Workers production profiling is useful, but its limits are easy to misread. You can capture a CPU profile or a heap-allocation profile from a deployed Worker or Durable Object for 1,000 to 50,000 milliseconds. Heap captures record one allocation stack trace per 512 kB allocated; Cloudflare does not publish a numeric CPU-profiler sampling interval, a profiler-overhead percentage, a capture-rate quota, or a retention period for profile files. A capture also requires a recently active, loaded isolate receiving traffic that exercises the code you want to inspect. Those constraints are separate from your Worker’s per-request CPU limit and from the seven-day windows documented for traces and logs.
What are Cloudflare Workers’ production profiling limits?
The production profiler has four practical boundaries: capture duration, live traffic, profile type, and service-side request limiting. The documented CLI and API duration is 1,000–50,000 milliseconds. That is the length of a profiling capture, not the maximum CPU time a Worker request may consume.
As an Amazon Associate I earn from qualifying purchases.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Profiling operates on a live deployment. The selected Worker version or Durable Object must already be loaded, and requests must execute the route or function of interest while the capture is running. Starting a capture does not replay historical requests, invoke your code, or turn an old log range into a retrospective profile. If the isolate is idle, the resulting file can contain little or no useful activity.
Cloudflare documents both CPU and heap-allocation captures. CPU profiles help locate functions consuming CPU time. Heap profiles show where allocations originate during the window; they are not snapshots of objects that remain in memory. The production guide states: “A Heap profile measures allocations, not retained memory.”
#1 Best Overall
- [Ultra-Lightweight Ear Buds Designed for Small Ears] Each earbud weighs only 3.7g and features a compact, ergonomic in-ear design made especially for small ears. Secure, low-profile, and comfortable for workouts, all-day wear or overnight sleeping.The ultra-slim profile sits flush in your ear with no hard edges poking or rubbing, so you can even sleep on your side without discomfort.
- [Immersive Stereo Sound with TOZO OrigX Technology] TOZO OrigX tuning delivers clear vocals, balanced mids, and natural stereo sound for music, podcasts, and videos.
- [Long Battery Life for Daily Use] Get up to 7 hours of playtime on a single charge, with up to 32 hours total using the charging case—ideal for workdays, commuting, and extended listening sessions.
- [Bluetooth 5.3 & Stable Connection] Bluetooth 5.3 provides fast pairing, stable wireless performance, and reduced dropouts as you move around home or office.
- [Deep Bass with Clear Vocals] High-performance drivers produce punchy bass while keeping vocals clean and detailed for everyday listening.
There is also a service-side rate limit. An API request can receive HTTP 429; when Cloudflare supplies a Retry-After header, wait for that interval before trying again. The reviewed documentation does not publish a numeric quota, so do not design an automation schedule around an assumed requests-per-minute value. Cloudflare’s production profiling guide is the authoritative reference for these capture rules.
How do CPU and heap profiles differ?
Choose the profile type according to the signal you need. A CPU profile answers “Which functions are using execution time?” A heap profile answers “Which call stacks are responsible for allocations during this capture?” Neither profile answers every memory question, and a heap-allocation profile cannot prove that allocated objects are still live.
| Profile | Best for | What it records | What it cannot establish |
|---|---|---|---|
| CPU | Finding hot functions, expensive parsing, loops, or unusually costly request paths | CPU-time activity during the live capture | The reviewed production guide does not state a numeric sampling interval or an overhead percentage |
| Heap allocation | Finding code paths that allocate heavily | One allocation stack trace for each 512 kB allocated during capture | Whether those allocations remain retained or represent a leak |
Use CPU first when latency or CPU consumption is the symptom. Use heap when garbage-collection pressure, allocation volume, or unexpectedly large temporary objects are the concern. If you need retained-object analysis, treat the heap profile as a lead generator and use an appropriate memory-analysis method rather than calling it a heap snapshot.
What is the Cloudflare Workers profiling sampling rate?
The documented heap cadence is one stack trace per 512 kB allocated during the capture. Cloudflare does not state a numeric sampling interval for production CPU profiles in the reviewed guide. Therefore, it is inaccurate to describe 512 kB as the CPU profiler’s rate or to invent a CPU interval from another observability product.
Heap sampling is allocation-based
The 512 kB figure is tied to bytes allocated, not elapsed time. A busy allocation path can produce many samples in a short period, while a mostly idle capture may produce few. Because the sample represents allocation stacks, it indicates allocation sources rather than the amount of memory ultimately retained.
CPU sampling is not numerically specified
Cloudflare’s production documentation describes CPU profiles as a way to identify functions consuming CPU time but does not publish a frequency such as “one sample every N milliseconds.” Keep that limitation visible when comparing profile output or building internal runbooks. A visual density in a pprof viewer is not a license to infer an undocumented interval.
Rank #2
- 【IP68 Waterproof Multifunctinal Smart Watch】Torjalph fitness activity tracker smart watch is IP68 waterproof technology ,you can wear it when swimming,handwashing,it's rain-proof which means that you perform multiple exercise modes outside without worrying about sweat or rain corroding your fitness smart watch,and this watch equipped with a 230 mAh large capacity battery magnetic charging ,and this sports activity tracker watch can be used 5-7 days If fully charged.
- 【Larger 1.83'' Full Screen Watch & Multi-dial Selection】1.83 inch full HD color touch screen smart watch for men / women provides excellent 240*284 resolution image quality ,This Bluetooth smart watch offers a better touch experience and remarkable immersion and ,more than 200 watch faces for you to choose from App"Da Fit ", you can change the dial style according to your mood to wear keeping the world on your wrist
- 【Accurate Activity Tracker Smart Watch 】sports watch activity tracker uses sweat and dust resistant breathable silicone strap, our Fitness Tracker Smart Watch for Android iOS Phones supports 12 sports modes,walking,running,bicycle,swimming,badminton,basketball,football,rope skipping. This bluetooth calls smart watch for men women is lightweight and stylish, bringing more convenient and relaxing sports experience while also providing more protection.
- 【Your 24 Hour Personal Daily Health Monitor】Fitness watch for women men full touch screen IP67 waterproof bluetooth smartwatch for Android iOS Phones with heart rate blood pressure monitor multifunction Andriod smart watch with steps counter sleep monitor sports fitness watch, monitoring blood oxygen saturation allows you to understand your physical condition in real time to track fitness progress ,to know your health status in time.
- 【More Functions Bluetooth Call Smart Watch】This fitness activity smart watch full screen for iPhone and Android phones includes a bunch of new features including:The bluetooth smart watch can hang up the phone, receive and display WhatsAPP /Text /Facebook /Twitter /Messenger reminder.Support sedentary reminder, alarm clock, stopwatch, timer,weather information,music control, remote camera control compatible with Android 5.0 and above or iOS 9.0 and above smartphones.
Tracing rates are a different control
Workers tracing uses head-based sampling, documented as a rate from 0 to 1 and defaulting to 1 when tracing is enabled without another value. Cloudflare says unsampled requests incur no tracing overhead. Those statements apply to tracing, not to an active CPU or heap-profile capture, so they cannot be used to claim that profiling has zero overhead or to supply a profiler sampling rate. See Workers tracing documentation for the separate tracing behavior.
Does production profiling add CPU or latency overhead?
Cloudflare does not publish a numeric CPU, wall-clock, or latency overhead figure for an active production profile capture in the profiling guide. The defensible operational position is to treat overhead as an unquantified cost, capture briefly, and watch normal service indicators while profiling. Do not promise “zero overhead,” and do not substitute tracing’s unsampled-request statement for a profiler measurement.
A short capture reduces exposure, but it does not eliminate the need for change control. Profile during a representative traffic period, avoid combining it with unrelated load tests, and compare request CPU and latency metrics with a nearby period when no capture was active. Those comparisons are an operational safeguard, not a Cloudflare-published benchmark.
How long can a profile capture run?
The documented CLI and API accept a duration from 1,000 ms (one second) through 50,000 ms (50 seconds). This is a profiler window. It does not extend, reduce, or otherwise modify the execution limit applied to an individual Worker request.
Do not confuse capture duration with request CPU limits
Cloudflare’s platform limits are separate:
| Workers plan | Per-request CPU limit | Important qualification |
|---|---|---|
| Free | 10 ms for an HTTP request | CPU time spent executing Worker code; waiting for network responses does not count |
| Paid | 30 seconds by default | Can be raised to a maximum of 5 minutes (300,000 ms) |
Cloudflare reports that the average Worker uses approximately 2.2 ms of CPU per request, while heavier authentication, server-side-rendering, or large-payload workloads commonly use 10–20 ms. These are published workload figures, not measurements of your profile or a guarantee for any application. Consult Workers limits when checking the execution ceiling for your plan.
What must be true before you start a production capture?
A successful request is not enough by itself; the right isolate must be active and the right code must execute during the window. Check these conditions first:
Rank #3
- [Low Noise High Fidelity] New bee guitar cord with 22AWG oxygen-free-copper center conductors provides enhanced signal transfer, the PVC insulation material and the high-density stranded copper double shield provide noise rejection for quiet operation
- [Ideal for Choice] New bee 1/4 straight and right angle metal connectors, ensures a secure connection in any instrument, pedal, or amp jack. L shape great fro on stage guitar stands
- [Wide Compatible] Suitable for mixer, power amplifier, microphone, effector, decoder, equalizer, electoronic organ, electric guitar, electric bass and other electronic equipments. Widely used in the installation of smart home, karaoke speaker, professional stage audio, etc
- [Easy Disassemble] Making incidental repairs quick and easy! Very nice, flexible cable
- [Dont Hesite to Order] 24 hours customer services and professional technology team are standing by
- Deployment: select the deployed Worker version or Durable Object instance you intend to study.
- Recent activity: the isolate must be loaded and recently active rather than cold and unused.
- Representative traffic: send real or controlled requests that reach the route, binding, or function under investigation.
- Readable symbols: enable source maps before capture if you need source-file names and locations instead of bundled output.
- Access: API use requires a token with Workers Scripts Read permission or the Content Read-Only Workers role.
Profiling does not generate traffic. A capture taken while a health check hits a different route will faithfully profile the health check, not the expensive endpoint you had in mind.
How do you capture and inspect a profile?
Cloudflare supports the dashboard, its CLI, and an API. The controls differ by interface, but the operational sequence is the same.
- Define the question. Decide whether you are looking for CPU-consuming functions or allocation sources. Select CPU or heap accordingly.
- Prepare symbols. Turn on source-map upload before deploying the version you will profile. Wrangler uploads source maps by default starting with version 4.150.0, according to the current production guide; verify your deployment process if you use an older version or a custom build.
- Confirm access and target. In the dashboard, CLI, or API, choose the Worker version or Durable Object target and make sure your account or API token has the required read permission.
- Set a bounded duration. Use a value between 1,000 and 50,000 ms. Start with the shortest window that is likely to include representative requests, then extend only when the workload is too sparse.
- Exercise the code. During the capture, route representative requests through the suspected handler, parser, binding, or Durable Object method. Avoid relying on requests that occurred before the start time.
- Download the artifact. Profiles captured through the dashboard, CLI, or API can be downloaded and opened with pprof-compatible tools.
- Correlate before changing code. Compare the profile’s hot functions or allocation stacks with request logs, CPU metrics, and the deployed source revision. A profile is a sample of one live window, not proof that every request behaves identically.
The API may return 429 Too Many Requests. Wait and honor Retry-After when present instead of immediately retrying in a tight loop. If the capture contains no useful samples, first check traffic and isolate activity before changing the duration.
How long does Cloudflare retain Workers profiles?
The public production-profiling guide does not state a retention period for captured CPU or heap profile files. You should therefore download artifacts you need for an investigation and store them under your organization’s retention and access rules; do not promise that a profile will remain downloadable for seven days or any other fixed period.
Cloudflare documents retention for related products, but those windows cannot be transferred to profile artifacts:
| Data type | Documented window | Can it be used as profile retention? |
|---|---|---|
| Workers traces | Seven days | No; traces are a separate observability dataset |
| Workers Logs | Maximum seven days | No; log retention does not define profile-file retention |
| Worker metrics | Inspectable up to three months back, in increments of up to one week | No; metrics lookback is not an artifact-retention promise |
Read the separate documentation for traces, Workers Logs, and metrics and analytics when planning those datasets. None supplies the missing profile-retention figure.
Rank #4
- Digital Family Calendar: BSIMB digital calendar wall touch screen puts an end to your chaotic life. Designed for busy families, the digital calendar keeps you organised and in step with your family. Assign different colours and tasks to each family member to make family scheduling easier and more intuitive.
- Easy Setup and Auto-Sync: Plug in, connect to WLAN and auto-syn your electric calendar in minutes. Digital wall calendar compatible with Google calendar, iCloud calendar, Outlook calendar, Cozi calendar and Yahoo calendar. Add events directly on the device or through the ecalendar app. Electronic calendar stays current whether you're at work or traveling—no manual updates needed.
- All-In-One Home Assistant: Smart calendar builds habits and encourage independence with personalised chores. Display daily meal plan to reduce the daily ‘What's for dinner?’ stress. To-do lists keep track of everything you want. Turning chores into valuable achievements with stars and rewards, making life easier and more enjoyable.
- Manage Anytime Anywhere: BSIMB electronic calendar for family stay connected with the powerful ecalendar app. Access and manage your digital wall calendar anytime, anywhere with free features like stars & rewards, photo screensaver, chore charts, meals plans, to-do lists etc. Electronic planner paid features like Magic Import tailor advanced tools for users with special needs.
- Stay in Touch with Loved Ones: Share photos or videos from your phone to electronic wall calendar with app, instantly update your life moment, and keep in touch with your family, especially those who are far away; you can also surprise your smart calendar digital wall by sending him a wish card; it's also great for keeping tech-illiterate elders in touch with the youngsters.
What do source maps and pprof add to the workflow?
Without source maps, a production bundle can make a profile difficult to read: names may point to generated files, minified functions, or wrapper code. Uploading source maps with the deployment lets the profiling interface and downloaded data map samples back to meaningful source locations when the toolchain and revision match.
After downloading, use a pprof-compatible viewer to inspect CPU hotspots or allocation call stacks. Preserve the profile together with the Worker version, deployment timestamp, source revision, capture type, and duration. That metadata prevents a later analyst from mistaking a heap-allocation sample for a retained-memory snapshot or comparing profiles from unrelated builds.
What should you do when a capture is empty or misleading?
- No samples: verify that the selected version or Durable Object was loaded and that traffic reached the code during the capture. Historical logs cannot fill an inactive window.
- Wrong functions: check the URL, route, feature flag, and bindings used by the requests. The profiler records what actually ran, including a low-cost health endpoint if that is what received traffic.
- Unreadable names: deploy with source maps and repeat the capture against the matching revision.
- Too little allocation detail: a short or lightly loaded heap window may not cross many 512 kB allocation intervals. Increase representative traffic or use a longer duration within the 50,000 ms maximum.
- HTTP 429: stop issuing retries, wait, and follow
Retry-Afterif Cloudflare returns it. The numeric capture quota is not published in the production guide. - Unexpected CPU impact: stop the capture, compare service telemetry with a non-profiled period, and treat the overhead as unknown rather than assuming tracing guidance applies.
Can you profile a Worker locally instead?
Yes. Cloudflare documents CPU profiling through Wrangler DevTools. Local reproduction can be useful for a repeatable code path, but the documentation warns that matching production behavior can be difficult. Use requests that resemble production and, where appropriate, production-like data through remote bindings so that the local profile exercises comparable work.
A local profile should not be presented as a production-limit measurement. Runtime configuration, traffic shape, data size, bindings, and isolate state can all differ. Use local DevTools to narrow a suspected function, then confirm the finding with a bounded production capture when the operational risk is acceptable. The local CPU-profiling guide is at Cloudflare’s CPU-usage documentation.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsHow should profiling fit with tracing, logs, and metrics?
Use each signal for a different question. Profiles explain where CPU time or allocations originate inside a live execution. Traces show request paths and spans, logs provide event detail, and metrics reveal aggregate behavior over time. Sampling controls for traces and logs should be configured separately from profile captures.
For a focused incident, capture a profile while a trace or log sample identifies the affected route, then correlate timestamps and deployment revision. Do not infer profile retention from trace or log retention, and do not call a trace sampling rate a CPU-profile sampling interval. This separation prevents an observability setting from being mistaken for a runtime or profiling limit.
Which profile should you choose?
| If your question is… | Choose | Plan the capture around… |
|---|---|---|
| “Which function consumes CPU?” | CPU profile | A route and traffic pattern that reproduce the slow or expensive work |
| “Where are allocations coming from?” | Heap-allocation profile | Enough representative allocation activity to produce samples; interpret results as allocation stacks |
| “Which requests are affected over hours or days?” | Traces, logs, and metrics | Their own sampling and retention settings, not a profile capture |
| “What objects remain live?” | Not answered by the documented heap profile alone | A separate retained-memory investigation |
The most reliable workflow is narrow and explicit: define the signal, deploy readable symbols, target an active isolate, drive representative traffic for 1–50 seconds, download the artifact, and record the exact revision and time window. That approach respects the documented limits without filling the gaps with assumptions.
Frequently Asked Questions
Can a Cloudflare Workers profile include requests that happened before I started it?
No. A production capture observes the selected loaded isolate only during its active window. Earlier requests remain available only through whatever logs, traces, or metrics you retained separately.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Why does my downloaded profile show bundled function names instead of my source files?
The deployed revision may not have matching source maps, or the maps may not have been uploaded for that build. Deploy the same revision with source maps enabled, then capture again so symbol locations can be resolved consistently.
Is a 512 kB heap sample the same thing as a 512 kB memory limit?
No. It is an allocation-sampling cadence: a stack trace is recorded after each 512 kB allocated during the capture. It neither caps memory nor reports how many bytes remain live.
What should an automated client do after a profiling API 429 response?
Stop immediate retries, wait for the interval in the Retry-After header when one is returned, and then retry with backoff. Cloudflare’s guide does not publish a fixed numeric profiling quota.
Can local Wrangler DevTools reproduce production profiling exactly?
Not necessarily. Differences in traffic, bindings, data, and isolate state can change the result; use production-like requests and remote bindings where suitable, then validate important findings with a controlled production capture.
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
