If a visitor saw an error but the website worked a minute later, record each occurrence right away. You need the URL, time, action, and response. A few precise entries can help your hosting provider or developer find a recurring condition.
reChecker checks availability on a five-minute schedule, so a brief problem may occur between checks. Don’t dismiss a report just because the monitoring result was successful; compare it with other data.
Separate different reports into different cases
“The website sometimes goes down” might mean that a page won’t open, takes a long time to load, or has an unresponsive button. Ask the person to describe the action and the response they saw. Don’t combine a form submission error with a failure to open the home page.
Save the exact URL, including any important parameters. Find out whether the person was signed in and what device and connection they used. If this information isn’t available, leave the field blank; guessing the phone model won’t help with the investigation.
Ask people to note the time the next time it happens, rather than reconstructing it from memory in the evening. An approximate time is also useful if it’s labeled as approximate. For example: “around 15:00, after receiving an email.” Don’t turn this note into an exact outage start time.
Fill the log with facts, not conclusions
You can keep the log in a regular team spreadsheet. It’s your document for investigating the issue, not a separate built-in reChecker feature. Here’s an illustrative example:
| Time, MSK | URL and action | Conditions | Result |
|---|---|---|---|
| 04.10, 10:03 | Open /catalog | Phone, mobile network | Error; opened at 10:04 |
| 04.10, 12:18 | Open /catalog | Computer, office network | Same error; retry at 12:19 succeeds |
| 05.10, 10:02 | Open /catalog | Computer, office network | Error; screenshot saved |
Three entries let you check one URL across several intervals. They don’t prove that the entire website was down or establish a common cause.
Don’t copy all visitor correspondence or order data into the spreadsheet. To start, you need the observed response and the conditions. Assign someone to collect new entries and pass them to the person responsible; otherwise, the log may stop after the first day.
Compare incidents with checks and successful attempts
In the “Site Monitoring” dashboard, look at the nearest actual availability checks. Record the time and result. Successful responses before and after a report don’t mean the problem couldn’t have happened: this is illustrated on the five-minute timeline.
If there’s no data for the relevant interval, don’t fill it in with “the website was working.” If both monitoring and a manual attempt captured the error, save both sources. Check whether they tested the same URL.
Also look for a successful attempt under similar conditions. Differences can be useful: the same URL opens on a computer but not on a phone; the failure occurs only after signing in. These point to what to investigate, but don’t explain the cause by themselves. If the issue recurs around the same time, compare that timing with the logs first.
Ask for specific intervals to be checked
Example request: “For /catalog, errors were recorded at 04.10 at 10:03 and 12:18 MSK, and at 05.10 at 10:02. A retry a minute later succeeded. The devices and networks are listed in the table. Please check the server logs for these intervals and identify a common condition or tell us what additional data is needed.”
Don’t just ask them to “restart everything”: the symptom disappearing doesn’t explain the cause. Agree on the next deliverable—a review of the relevant records, a difference found, or a specific way to reproduce the issue. The request fields are in the developer task template.
After the fix, repeat the same actions under the same conditions. Continue the log, noting when the change was published. One quiet hour doesn’t prove that a rare issue has been permanently resolved. If it happens again, a new entry after the fix will help compare it with earlier ones.