<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Andrew Molyuk</title>
    <description>The latest articles on DEV Community by Andrew Molyuk (@andrewmolyuk).</description>
    <link>https://dev.to/andrewmolyuk</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4210480%2F79c5971e-d630-4b2f-b680-a0d8ad7e0127.png</url>
      <title>DEV Community: Andrew Molyuk</title>
      <link>https://dev.to/andrewmolyuk</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/andrewmolyuk"/>
    <language>en</language>
    <item>
      <title>Why Domains Expire Silently, Not From Negligence</title>
      <dc:creator>Andrew Molyuk</dc:creator>
      <pubDate>Thu, 19 Nov 2026 17:48:49 +0000</pubDate>
      <link>https://dev.to/andrewmolyuk/why-domains-expire-silently-not-from-negligence-2i93</link>
      <guid>https://dev.to/andrewmolyuk/why-domains-expire-silently-not-from-negligence-2i93</guid>
      <description>&lt;p&gt;If you manage other people's sites, you've either had this happen or you know someone it happened to: the domain just stops working. No warning, no error in your deploy pipeline, no monitoring alert — because most monitoring watches the site, not the thing the site's address depends on. The first sign is usually a client's Slack message, and it's rarely polite.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why domains actually expire
&lt;/h2&gt;

&lt;p&gt;It's almost never someone deciding not to renew. It's a chain of small, boring failures, any one of which is enough on its own: the card on file at the registrar expired eight months ago and nobody noticed because autorenew failed silently instead of loudly. The WHOIS/registrant contact email is an old employee's inbox, or the client's own address that they never check, or — worse — an email address that itself lives on the domain that's about to expire, so the renewal notice can't even be delivered. ICANN requires registrars to send expiration notices, but a required email and a read email are different things; spam filters exist, inboxes get abandoned, forwarding rules break.&lt;/p&gt;

&lt;p&gt;None of this is a competence problem. It's a coverage gap — the renewal reminder depends entirely on channels the registrar controls and you don't: their email deliverability, their timing, whichever contact address happened to be on file when the domain was registered, possibly years before you ever touched the project.&lt;/p&gt;

&lt;h2&gt;
  
  
  The math nobody does until it's too late
&lt;/h2&gt;

&lt;p&gt;A domain costs somewhere around $10–15 a year. That's such a small number that losing track of it feels absurd in hindsight — but the cost was never the renewal fee, it's what happens in the gap. DNS resolution for a suspended domain can break within hours of expiry, taking the site, and if it's hosted there, email, down with it. If nobody catches it during the registrar's grace or redemption period, the domain becomes available again — and anyone, including a squatter who specifically watches for expiring domains with existing backlinks and traffic, can register it out from under you. Getting it back at that point isn't a renewal anymore; it's a negotiation, if it's possible at all.&lt;/p&gt;

&lt;h2&gt;
  
  
  What actually catches this
&lt;/h2&gt;

&lt;p&gt;Not the registrar's own reminder — that's the exact channel that already failed to reach anyone, or this wouldn't be a problem. Not a calendar reminder either, past two or three domains; it doesn't scale, and it's yet another thing that depends on someone remembering to set it up correctly in the first place. What actually works is checking the registration data independently, on a schedule, from a source that doesn't route through the registrant's own inbox.&lt;/p&gt;

&lt;p&gt;That's what a domain monitor is for. Checkmeup's does a straightforward RDAP lookup — the structured, standardized successor to WHOIS — once a day for every domain you add, and compares the registration's actual expiry date against three thresholds: 30 days out, 14 days out, and 7 days out, plus a final alert if it actually expires. Each threshold only fires once, so renewing resets it cleanly instead of leaving a stale warning around or requiring you to manually dismiss anything. Alerts go out over whichever channels you've already got wired up for everything else — Telegram, Slack, email, and SMS on paid plans — so a domain nobody's watching becomes a domain that pings you the same way a downed cron job or an expiring SSL cert does.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;&lt;a href="https://checkmeup.net/tools/domain-expiry-checker" rel="noopener noreferrer"&gt;Check a domain's expiration date now with the free domain expiry checker&lt;/a&gt;&lt;/strong&gt;&lt;/p&gt;

&lt;h2&gt;
  
  
  A short checklist, if you're doing this by hand today
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Put your own email as the technical/admin contact on client domains, not the client's — you're the one who needs the notice, and their inbox is one more point of failure you don't control.&lt;/li&gt;
&lt;li&gt;Keep one password-manager entry per domain with the registrar, expiry date, and renewal method — a five-minute setup cost that turns "which of our 40 domains needs attention" into a search.&lt;/li&gt;
&lt;li&gt;Turn on the registrar's autorenew and monitor independently — autorenew fails silently often enough (expired card, changed billing email) that it shouldn't be the only layer.&lt;/li&gt;
&lt;li&gt;Review the full domain list on a quarterly cadence, not just when something breaks — cheap insurance against the ones nobody's actively thinking about.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;None of this is exotic. It's the same instinct that makes uptime monitoring obvious — you wouldn't wait for a client to tell you their site is down — applied to the one dependency that sits underneath the site itself.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;— Andrew&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Originally published on the &lt;a href="https://checkmeup.net/blog/domain-expiry-monitoring" rel="noopener noreferrer"&gt;Checkmeup blog&lt;/a&gt;. This article was written with the help of AI (Claude) and reviewed by the author.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>dns</category>
      <category>webdev</category>
      <category>devops</category>
      <category>monitoring</category>
    </item>
    <item>
      <title>Cron Job Monitoring: Catch Silent Failures</title>
      <dc:creator>Andrew Molyuk</dc:creator>
      <pubDate>Sun, 08 Nov 2026 20:52:33 +0000</pubDate>
      <link>https://dev.to/andrewmolyuk/cron-job-monitoring-catch-silent-failures-5122</link>
      <guid>https://dev.to/andrewmolyuk/cron-job-monitoring-catch-silent-failures-5122</guid>
      <description>&lt;p&gt;A cron job doesn't usually announce its own death. It doesn't throw an exception into your error tracker, it doesn't page anyone, it doesn't even exit non-zero half the time — it just stops appearing in the log, and nothing downstream is looking for its absence. The backup that quietly stopped running three weeks ago, the invoice job that's been skipping every other client since a dependency bump, the cleanup script that died the day the server rebooted and the crontab didn't survive it — these are the failures that don't show up until someone asks for a backup that isn't there.&lt;/p&gt;

&lt;h2&gt;
  
  
  What is cron job monitoring?
&lt;/h2&gt;

&lt;p&gt;Cron job monitoring answers a question your crontab can't: did this job actually run? It works as a dead man's switch, inverted from how most monitoring works. An uptime check calls your server and expects an answer; a cron monitor waits for your job to call it, and raises an alarm if the call doesn't come. The job itself reports success — the monitor's only role is noticing when a report was expected and didn't arrive.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why cron jobs fail silently
&lt;/h2&gt;

&lt;p&gt;Cron itself has no concept of "this should have run" — it fires the schedule and moves on, whether or not the job existed to receive it. A handful of ordinary events can break a job without ever producing an error anyone sees:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;A server reboot (patching, a host migration) and the crontab entry never made it back, or came back with the wrong PATH&lt;/li&gt;
&lt;li&gt;A dependency, API key, or credential the script relies on expired or changed, and the script exits early — sometimes with status 0&lt;/li&gt;
&lt;li&gt;cron's own mail delivery is unconfigured or going to an inbox nobody reads, so a script that does error correctly is still invisible&lt;/li&gt;
&lt;li&gt;The previous run is still stuck (mid-download, waiting on a lock) and the schedule just skips the next one&lt;/li&gt;
&lt;li&gt;A deploy overwrote or commented out the crontab line and nobody noticed because nothing broke immediately&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;None of these look like an incident from the server's point of view. The only way to catch them is to expect a signal on a schedule and notice when it stops arriving — which is exactly what a cron monitor is built to do, and cron itself is not.&lt;/p&gt;

&lt;h2&gt;
  
  
  Setting up cron job monitoring in three steps
&lt;/h2&gt;

&lt;h3&gt;
  
  
  1. Pick a grace period that matches how the job actually behaves
&lt;/h3&gt;

&lt;p&gt;A grace period is how long the monitor waits past the expected time before calling a run missed. Too tight, and normal jitter — a slow disk, a busy host, a job that runs 90 seconds long instead of 60 — trips a false alarm and trains you to ignore alerts. Too loose, and a real failure sits unnoticed for hours. Base it on the job's own variance: watch a few real runs first if you don't already know how long it takes end to end, and set the grace period a bit above the slowest one you've seen, not the average.&lt;/p&gt;

&lt;h3&gt;
  
  
  2. Call a URL at the end of the job, not the start
&lt;/h3&gt;

&lt;p&gt;The monitor only knows a job succeeded if the job tells it so — chain the ping onto the command that already runs, after the real work, so a failing script never gets the chance to report success it didn't earn:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;&lt;span class="c"&gt;# in your crontab or script&lt;/span&gt;
0 3 * * * /opt/scripts/nightly-backup.sh &lt;span class="o"&gt;&amp;amp;&amp;amp;&lt;/span&gt; curl -fsS https://checkmeup.net/ping/YOUR_TOKEN
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;&lt;code&gt;&amp;amp;&amp;amp;&lt;/code&gt; is doing the important work here: curl only fires if the backup script exited 0. A script that dies partway through never sends the ping, which is exactly the failure the monitor exists to catch. &lt;code&gt;-f&lt;/code&gt; also makes curl itself fail loudly on a bad response instead of silently swallowing it.&lt;/p&gt;

&lt;h3&gt;
  
  
  3. Wire up an alert channel you'll actually see in time
&lt;/h3&gt;

&lt;p&gt;An alert that lands in an inbox nobody checks until Monday isn't meaningfully different from no alert at all. Pick something you'll notice within the grace period that actually matters for that job — a chat app you already have open, an SMS for anything client-facing, a webhook into whatever on-call system you already run. The channel matters as much as the check.&lt;/p&gt;

&lt;h2&gt;
  
  
  What to check before trusting a tool with it
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;An unguessable ping URL — it's the only credential protecting the endpoint, so it needs to not be a sequential ID&lt;/li&gt;
&lt;li&gt;A per-job grace period, not one global timeout for every schedule you run&lt;/li&gt;
&lt;li&gt;An execution log you can actually look at — not just a current status badge, but a history of when pings did and did not arrive&lt;/li&gt;
&lt;li&gt;A monitoring endpoint that always returns 200, so a blip in the monitoring service can't fail a job that's actually fine&lt;/li&gt;
&lt;li&gt;More than one alert channel, so a single dead integration (an expired bot token, a bounced email) never becomes a silent monitoring gap on top of the silent job failure it was supposed to catch&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  The mistake that undoes all of this
&lt;/h2&gt;

&lt;p&gt;Setting up monitoring and never simulating a failure is the most common way this quietly stops working. Comment out the cron line for a minute, or send a bad exit code on purpose, and confirm the alert actually arrives, on the channel you expect, within the grace period you set. A monitor nobody has ever seen fire is a monitor nobody actually knows works — the same blind spot as the cron job it's meant to be watching.&lt;/p&gt;




&lt;p&gt;Checkmeup's cron monitor is exactly the dead man's switch described above — unique ping URL, per-job grace period, execution and incident logs, alerts over Telegram, email, SMS, or webhook. The Hobby plan is free for up to 10 monitors, no credit card required, if you want to try it on the job you're least sure is still running.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;— Andrew&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Originally published on the &lt;a href="https://checkmeup.net/blog/cron-job-monitoring-guide" rel="noopener noreferrer"&gt;Checkmeup blog&lt;/a&gt;. This article was written with the help of AI (Claude) and reviewed by the author.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>devops</category>
      <category>linux</category>
      <category>monitoring</category>
      <category>tutorial</category>
    </item>
    <item>
      <title>The 3 AM SSL Cert That Cost Us a Client</title>
      <dc:creator>Andrew Molyuk</dc:creator>
      <pubDate>Sat, 31 Oct 2026 12:06:07 +0000</pubDate>
      <link>https://dev.to/andrewmolyuk/the-3-am-ssl-cert-that-cost-us-a-client-5f71</link>
      <guid>https://dev.to/andrewmolyuk/the-3-am-ssl-cert-that-cost-us-a-client-5f71</guid>
      <description>&lt;p&gt;9:14 AM. A Slack message from a client, no greeting: "is the site down for you too??" Then a screenshot — not a 500 page, not a blank white screen, but a full-bleed red warning. NET::ERR_CERT_DATE_INVALID. Your connection is not private. Every browser on earth was refusing to even load the page, and had been since roughly 3 AM, which meant six hours of checkout traffic had been quietly bounced before a single human noticed.&lt;/p&gt;

&lt;h2&gt;
  
  
  The part that stings isn't the outage
&lt;/h2&gt;

&lt;p&gt;A server crash is embarrassing but forgivable — things break, that's infrastructure. An expired TLS certificate is different, and it feels different, because it was never a mystery. Certificate lifetimes aren't random; they're printed in the cert itself the day it's issued. Nothing failed unpredictably. A date on a calendar just arrived, and nobody was watching the calendar.&lt;/p&gt;

&lt;p&gt;That's the part that actually costs you sleep afterward — not the six hours of downtime, but replaying how avoidable it was. No exotic infrastructure failure, no DDoS, no upstream provider outage to point at. Just a 90-day Let's Encrypt cert, an auto-renewal cron job that had been silently failing since a server migration two months earlier, and nobody positioned to notice until a customer did.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why the failure is invisible until it isn't
&lt;/h2&gt;

&lt;p&gt;Right up until the expiry date, an SSL certificate gives you zero signal that anything's wrong. The site loads fine. The padlock shows fine. curl works fine. There's no gradual degradation to notice, no error rate creeping up in a dashboard somewhere — because the certificate itself doesn't care what day it is, right up until the exact second it does. Then every browser, all at once, independently decides your site is unsafe and refuses to render it. Not a slowdown. A cliff.&lt;/p&gt;

&lt;p&gt;And the automation that's supposed to prevent this is exactly the kind of thing that fails silently, too. certbot renew running as a cron job with no monitoring on the cron job itself is a single point of failure disguised as a safety net — it works until a permissions change, a path that moved during a migration, a webhook that silently started returning a non-200, or a renewal rate limit nobody expected. When it fails, it doesn't send an email. It just quietly stops, and the countdown to that cliff keeps running.&lt;/p&gt;

&lt;h2&gt;
  
  
  What it actually costs
&lt;/h2&gt;

&lt;p&gt;Six hours doesn't sound like much until you multiply it by every visitor who hit a full-screen security warning instead of a product page. Some closed the tab immediately — a security warning reads as "this site got hacked," not "routine cert hiccup," to anyone who isn't in this industry. Some searched the client's name plus "down" and found nothing reassuring. A few emailed support, which is how the client found out before the fix was even in motion. The revenue lost that morning was real, but the trust cost was worse and lasted longer — the client's next question, reasonably, was why they'd found out from their own customers instead of from the agency managing their infrastructure.&lt;/p&gt;

&lt;h2&gt;
  
  
  What actually catches this
&lt;/h2&gt;

&lt;p&gt;Not the renewal cron job's own success — that's the exact mechanism that already failed silently. What catches it is something external, checking the live certificate itself rather than trusting that the process meant to renew it worked. That's the whole idea behind Checkmeup's SSL monitor: it connects to your domain daily, reads the certificate that's actually being served, and alerts at 30, 14, and 7 days before expiry — plus a final alert if it lapses anyway. It doesn't care whether certbot ran, whether the renewal webhook fired, or whether anyone remembered to check. It only cares what's actually on the wire, which is the one thing that matters at 9:14 AM.&lt;/p&gt;

&lt;h2&gt;
  
  
  A short checklist, if you're carrying this risk today
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Monitor the certificate independently of the renewal process — a passing cron job and a valid certificate are not the same fact, and only one of them is what your visitors' browsers check.&lt;/li&gt;
&lt;li&gt;Set alert thresholds with real runway — 7 days before expiry is enough time to fix a broken renewal script; 7 hours is not.&lt;/li&gt;
&lt;li&gt;Confirm renewal actually happened after every server migration, DNS change, or webhook update — it's exactly the kind of automation that keeps working in tests and quietly stops in production.&lt;/li&gt;
&lt;li&gt;Make sure the alert reaches a person, not just a log file nobody tails — Telegram, Slack, email, SMS, whatever's actually checked at 3 AM.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;None of this is exotic — it's the same logic as watching a cron job's heartbeat instead of trusting the job ran, applied to the certificate underneath every request your site serves. The difference between a non-event and a 9:14 AM Slack message is usually just: something was watching before the browser had to.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;— Andrew&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Want to see where your own certificate stands right now? The &lt;a href="https://checkmeup.net/tools/ssl-checker" rel="noopener noreferrer"&gt;free SSL checker&lt;/a&gt; shows its expiry date, days left and issuer, no sign-up. Originally published on the &lt;a href="https://checkmeup.net/blog/ssl-certificate-expired-story" rel="noopener noreferrer"&gt;Checkmeup blog&lt;/a&gt;. This article was written with the help of AI (Claude) and reviewed by the author.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>ssl</category>
      <category>devops</category>
      <category>webdev</category>
      <category>monitoring</category>
    </item>
    <item>
      <title>RDAP vs WHOIS: Why Some Domains Hide Their Expiry</title>
      <dc:creator>Andrew Molyuk</dc:creator>
      <pubDate>Tue, 27 Oct 2026 09:51:13 +0000</pubDate>
      <link>https://dev.to/andrewmolyuk/rdap-vs-whois-why-some-domains-hide-their-expiry-2mci</link>
      <guid>https://dev.to/andrewmolyuk/rdap-vs-whois-why-some-domains-hide-their-expiry-2mci</guid>
      <description>&lt;p&gt;Ask a domain monitoring tool when example.com expires and you get a date in a second. Ask about an .io, .co.il or .de domain and, depending on the tool, you might get an error, a blank field, or a confident "domain not found" for a domain that's clearly live. That isn't the tool being flaky. Registration data for different top-level domains lives behind different protocols, and some registries don't publish an expiry date at all.&lt;/p&gt;

&lt;h2&gt;
  
  
  Where a domain's expiry date comes from
&lt;/h2&gt;

&lt;p&gt;Every top-level domain is run by a registry: Verisign for .com, ISOC-IL for .il, DENIC for .de. The registry is the source of truth for when a registration ends. There are two ways to ask it.&lt;/p&gt;

&lt;p&gt;RDAP (Registration Data Access Protocol) is the modern one: an HTTPS API that returns JSON, with the expiry as an event named "expiration". IANA keeps a bootstrap file that maps each TLD to its registry's RDAP server, and services like rdap.org use it to forward a query to the right place. Registries for generic TLDs like .com, .net, .org and the newer ones such as .app or .dev are required by their contracts with ICANN to run RDAP, so for those the answer is always one request away.&lt;/p&gt;

&lt;p&gt;WHOIS is the old one: you open a TCP connection to port 43, send the domain name, and read back plain text. There's no standard format. Every registry writes its own.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why some TLDs have no RDAP
&lt;/h2&gt;

&lt;p&gt;Country-code TLDs (.il, .it, .jp, .de and the rest) aren't bound by ICANN's registry contracts, so whether they run RDAP is up to each registry. Many do: .uk, .nl, .fr, .ca and .au all answer RDAP. A good number don't, including some you'd expect to, like .io, .co and .me, which are country codes even though most people use them as generic names.&lt;/p&gt;

&lt;p&gt;For a domain under one of those, an RDAP lookup through a bootstrap redirector fails with a 404, exactly the response it gives for a domain nobody registered. A tool that reads that 404 as "not registered" will tell you your live domain doesn't exist.&lt;/p&gt;

&lt;h2&gt;
  
  
  What WHOIS looks like when there is no RDAP
&lt;/h2&gt;

&lt;p&gt;The registries without RDAP still answer WHOIS, each in its own shape. The expiry for an Israeli domain looks like this:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight plaintext"&gt;&lt;code&gt;domain:       isoc.org.il
assigned:     11-01-1996
validity:     11-01-2029
registrar name: Israel Internet Association ISOC-IL
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Day first, labeled "validity". Russia's registry says "paid-till: 2027-09-30T21:00:00Z". Italy's says "Expire Date: 2026-12-31". Japan's answers in Japanese unless you append "/e" to the query, and second-level names like .co.jp put the date inside a status line: "[State] Connected (2027/03/31)". The .io, .co, .me and .us registries happen to use the same layout as gTLD WHOIS, "Registry Expiry Date: 2027-05-01T00:00:02Z".&lt;/p&gt;

&lt;p&gt;So reading an expiry date over WHOIS means knowing each registry's format. A parser that just looks for anything date-shaped near the word "expir" will eventually read the wrong line, or read 01-11 as January 11 when the registry meant November 1. For a monitor that's worse than failing: a wrong date can tell you a domain is fine weeks after it lapsed.&lt;/p&gt;

&lt;h2&gt;
  
  
  The registries that publish no expiry at all
&lt;/h2&gt;

&lt;p&gt;Some registries don't put an expiry date in WHOIS. DENIC (.de), EURid (.eu) and nic.at answer WHOIS but leave the date out; DENIC's whole answer for denic.de is two lines, the domain and "Status: connect". For .de in particular there's nothing to leave out: a registration runs until it's cancelled, and whether it continues depends on your registrar being paid, not on a date at the registry. Others don't answer WHOIS for just anyone: SWITCH (.ch and .li) replies "Requests of this client are not permitted" and points to a web form, and Spain's .es registry only answers addresses on an allow list.&lt;/p&gt;

&lt;p&gt;For those domains, no external monitor can warn you about an expiry date, because there isn't one to read. The risk moves to your registrar account: the card on file, the renewal or billing emails, and who receives them.&lt;/p&gt;

&lt;h2&gt;
  
  
  How to check your own domain's expiry
&lt;/h2&gt;

&lt;p&gt;For a TLD with RDAP, one request does it; look for the "expiration" event in the response:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;curl -sL https://rdap.org/domain/example.com &lt;span class="o"&gt;|&lt;/span&gt; jq &lt;span class="s1"&gt;'.events[] | select(.eventAction == "expiration")'&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;For one without, ask its registry's WHOIS server directly. whois.iana.org tells you which server that is:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight shell"&gt;&lt;code&gt;whois -h whois.iana.org il &lt;span class="o"&gt;|&lt;/span&gt; grep ^whois     &lt;span class="c"&gt;# → whois.isoc.org.il&lt;/span&gt;
whois -h whois.isoc.org.il example.co.il &lt;span class="o"&gt;|&lt;/span&gt; grep validity
whois -h whois.jprs.jp &lt;span class="s2"&gt;"example.jp/e"&lt;/span&gt; &lt;span class="o"&gt;|&lt;/span&gt; grep &lt;span class="s2"&gt;"Expires on"&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;TLD&lt;/th&gt;
&lt;th&gt;Where the expiry comes from&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;.com, .net, .org, .app, .dev and other gTLDs&lt;/td&gt;
&lt;td&gt;RDAP&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.uk, .nl, .fr, .ca, .au, .in, .br and many ccTLDs&lt;/td&gt;
&lt;td&gt;RDAP&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.io, .co, .me, .us&lt;/td&gt;
&lt;td&gt;WHOIS, "Registry Expiry Date"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.il&lt;/td&gt;
&lt;td&gt;WHOIS, "validity" (DD-MM-YYYY)&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.it&lt;/td&gt;
&lt;td&gt;WHOIS, "Expire Date"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.ru&lt;/td&gt;
&lt;td&gt;WHOIS, "paid-till"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.jp&lt;/td&gt;
&lt;td&gt;WHOIS with /e, "[Expires on]" or "[State]"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.cn&lt;/td&gt;
&lt;td&gt;WHOIS, "Expiration Time"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.se&lt;/td&gt;
&lt;td&gt;WHOIS, "expires"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.de, .eu, .at&lt;/td&gt;
&lt;td&gt;Not published&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;.ch, .li, .es&lt;/td&gt;
&lt;td&gt;WHOIS only for allow-listed clients&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;That table is a snapshot from October 2026. Registries add RDAP over time, and IANA's bootstrap file at data.iana.org/rdap/dns.json is the current list.&lt;/p&gt;

&lt;h2&gt;
  
  
  How Checkmeup handles it
&lt;/h2&gt;

&lt;p&gt;Checkmeup's domain monitor uses RDAP wherever a registry runs it, and WHOIS for the TLDs in the table above that publish an expiry date there, with a separate parser for each registry, written against that registry's real answers. A domain the registry doesn't know shows as an error, never as expired. Adding a domain under a TLD whose expiry can't be read is refused with the reason, rather than accepted as a monitor that can only ever show an error. The reasoning behind those choices is in a &lt;a href="https://github.com/checkmeup/checkmeup/blob/main/.about/adr/domain-lookups.md" rel="noopener noreferrer"&gt;decision record on GitHub&lt;/a&gt;, and the full story of why domains lapse in the first place is in &lt;a href="https://checkmeup.net/blog/domain-expiry-monitoring" rel="noopener noreferrer"&gt;an earlier post&lt;/a&gt;.&lt;/p&gt;

&lt;p&gt;To look up one domain without the command line, the free &lt;a href="https://checkmeup.net/tools/domain-expiry-checker" rel="noopener noreferrer"&gt;domain expiry checker&lt;/a&gt; runs the same RDAP and WHOIS lookups, no sign-up.&lt;/p&gt;

&lt;p&gt;&lt;em&gt;— Andrew&lt;/em&gt;&lt;/p&gt;




&lt;p&gt;&lt;em&gt;Originally published on the &lt;a href="https://checkmeup.net/blog/rdap-vs-whois-domain-expiry" rel="noopener noreferrer"&gt;Checkmeup blog&lt;/a&gt;. This article was written with the help of AI (Claude) and reviewed by the author.&lt;/em&gt;&lt;/p&gt;

</description>
      <category>dns</category>
      <category>webdev</category>
      <category>devops</category>
      <category>networking</category>
    </item>
  </channel>
</rss>
