DNS_PROBE_FINISHED_BAD_CONFIG

DNS_PROBE_FINISHED_BAD_CONFIG is Chrome reporting the result of its own investigation. When a lookup fails, the browser runs a probe to work out why, and this verdict means it decided the DNS configuration on your device or network is at fault. Not that the domain is missing, but that the machinery meant to find it is not working.

Updated August 2026 · 6 min read

  • Written by

    Andrian Valeanu Andrian Valeanu Founder of Pulsetic

    Andrian Valeanu founded Pulsetic and, before it, Designmodo. Across 15-plus years he has shipped web products, design tools, and monitoring software teams around the world rely on.

  • Reviewed by

    Ionut Caval Ionut Caval Technical reviewer

    Ionut Caval reviews Pulsetic's technical guides for accuracy. He works hands-on with web servers, networking, and uptime monitoring day to day, and makes sure the causes and fixes here hold up in production.

The short version: Set your DNS servers explicitly to a public resolver such as 1.1.1.1 or 8.8.8.8, then flush the DNS cache and retry. This fixes the majority of cases and simultaneously proves the fault was the resolver configuration. If a VPN recently disconnected, reset the network adapter, since VPN clients often leave DNS settings pointing at a resolver that is no longer reachable.

Key takeaways

  • DNS_PROBE_FINISHED_BAD_CONFIG means Chrome ran its DNS probe and concluded the resolver configuration itself is broken, so no lookup could complete.
  • It is the opposite of NXDOMAIN. That error means DNS answered definitively that the name does not exist; this one means DNS never produced a usable answer at all.
  • The fault is local in almost every case: the device, the router, or the resolver they were told to use. The website is rarely involved.
  • Setting a public resolver such as 1.1.1.1 or 8.8.8.8 is both the usual fix and a diagnosis, because it proves the problem was the previous DNS configuration.
  • A VPN that exited badly is a common cause, leaving DNS settings pointing at a resolver only reachable through a tunnel that no longer exists.
Error type
DNS / browser error
Whose side
Almost always your device or network, not the website
Fix difficulty
Easy
Common cause
The DNS resolver configuration is not usable

How did you find out DNS broke?

Pulsetic is website uptime monitoring. It checks as often as every 30 seconds, confirms a DNS failure from another region, then emails you. Domain expiry is watched too.

10 monitors free, email alerts on failure and recovery. No credit card.

What does DNS_PROBE_FINISHED_BAD_CONFIG mean?

DNS_PROBE_FINISHED_BAD_CONFIG is not a raw network error so much as a conclusion. When a name fails to resolve, Chrome runs a diagnostic probe to establish why, and this message is one of its verdicts: the DNS configuration available to the device does not work. The resolver addresses may be wrong, they may point at a server that no longer answers, or the queries may be blocked before they arrive. What the browser is telling you is that it could not obtain a usable answer, which is a different situation from receiving an answer it did not like.

That distinction matters because it points the investigation firmly at your own side. NXDOMAIN, the other common probe verdict, means DNS worked exactly as designed and told you the name does not exist, which usually means a typo or a domain that has lapsed. BAD_CONFIG means the lookup never completed. Most devices receive their resolver addresses from the router over DHCP, so anything that corrupts that path breaks every lookup at once: a router handing out addresses that no longer work, a VPN client that exited without restoring what it changed, or security software intercepting DNS traffic. The rest of the connection keeps working, which is what makes the failure feel so strange.

YouDNSNetworkCDN / ProxyWeb serverApp / DB
The path a request takes from your browser to the website's servers. A DNS_PROBE_FINISHED_BAD_CONFIG is produced at the highlighted stages.
53
The UDP and TCP port DNS queries use, which some restrictive networks block outright
1.1.1.1
A public resolver worth setting explicitly, both as a fix and as a way to prove the fault was local
RFC 1035
The specification defining how DNS queries and responses work, published in 1987 and still current

How the DNS_PROBE_FINISHED_BAD_CONFIG error appears

The wording changes depending on your browser, device, or server. Here is how this error commonly shows up:

What a DNS_PROBE_FINISHED_BAD_CONFIG looks like in the browser. The exact wording varies by browser, device, and server.
  • DNS_PROBE_FINISHED_BAD_CONFIG
  • DNS probe finished bad config
  • This site can't be reached, DNS_PROBE_FINISHED_BAD_CONFIG
  • ERR_NAME_NOT_RESOLVED
  • DNS server not responding
  • Server IP address could not be found

The DNS probe verdicts, compared

Chrome reports several different outcomes from the same probe, and each one narrows the problem differently.

Result What it means Who fixes it
DNS_PROBE_FINISHED_BAD_CONFIG The resolver configuration is unusable, so no lookup could complete. You, by setting a working resolver or clearing VPN leftovers.
DNS_PROBE_FINISHED_NXDOMAIN DNS answered definitively that the name does not exist. You, by correcting the address, or the owner, by fixing their records.
DNS_PROBE_FINISHED_NO_INTERNET There is no working connection at all, so DNS was never the issue. You, by restoring the network connection first.
ERR_NAME_NOT_RESOLVED The name did not resolve, with no probe verdict attached. Either: your resolver, or the domain's DNS records.

What each DNS outcome tells you

Chrome distinguishes these because a lookup can fail in several ways, and each points somewhere different.

ResultWhat happenedWhere to look
DNS_PROBE_FINISHED_BAD_CONFIGNo usable answer; resolver settings look brokenYour device, router, or VPN leftovers
DNS_PROBE_FINISHED_NXDOMAINA definitive answer: the name does not existA typo, or the domain's own records
DNS_PROBE_FINISHED_NO_INTERNETThe probe found no working connection at allThe network itself, before DNS
ERR_NAME_NOT_RESOLVEDThe name did not resolve, without a probe verdictEither: resolver or records

What causes DNS_PROBE_FINISHED_BAD_CONFIG?

  • Resolver addresses handed out by the router that are wrong, unreachable, or pointing at a DNS server that has stopped answering.
  • A VPN client that exited badly and left DNS settings pointing at a resolver only reachable inside a tunnel that no longer exists.
  • A stale DNS cache on the device holding failed or outdated entries that keep being reused.
  • Security software or a firewall intercepting or blocking DNS traffic on port 53, so queries never reach a resolver.
  • A router in a bad state after an outage or a firmware update, still advertising DNS servers it can no longer reach.
  • A corrupted or leftover hosts file entry, which is consulted before DNS and can break one hostname while others resolve normally.
  • Chrome's own secure DNS setting pointing at a DNS-over-HTTPS provider that is unreachable on the current network, which restricted networks often block.

How to find the cause fast

  1. Open a raw IP address such as http://1.1.1.1 in the browser. Loading it while no hostname works confirms the fault is DNS and not the connection.
  2. Query a public resolver directly with dig @1.1.1.1 example.com. An answer there while the default resolver fails identifies the configuration as the problem.
  3. Check whether other devices on the same network resolve names. All failing points at the router; only one failing points at that device.
  4. Look at the adapter's current DNS servers, especially after a VPN disconnect, for addresses that no longer belong on that network.
What a DNS_PROBE_FINISHED_BAD_CONFIG looks like from the command line. The grey lines starting with # are explanatory comments.

How DNS_PROBE_FINISHED_BAD_CONFIG looks from the outside

DNS failures on a visitor's side never touch your infrastructure, so nothing about them appears in your logs. But the same class of problem has a version that is entirely yours: a nameserver that stops answering, a zone that fails to load, or records that lapse at the registrar. From inside your own network those failures are often invisible, because internal resolvers and caches keep answering from memory long after the authoritative source has stopped. Checking resolution from outside, from several networks, is the difference between finding out from your own alerting and finding out from the first person who could not reach you.

How to fix DNS_PROBE_FINISHED_BAD_CONFIG

If you are a visitor

  1. Set your DNS servers explicitly to a public resolver such as 1.1.1.1 and 1.0.0.1, or 8.8.8.8 and 8.8.4.4, then retry.
  2. Flush the DNS cache (ipconfig /flushdns on Windows, sudo dscacheutil -flushcache; sudo killall -HUP mDNSResponder on macOS).
  3. Renew the DHCP lease with ipconfig /release then ipconfig /renew, so the device requests resolver addresses again.
  4. Reset the network adapter after any VPN disconnect, since clients frequently leave their DNS settings behind when they exit badly.
  5. Restart the router if every device on the network is affected, and check whether it is still advertising DNS servers it can reach.
  6. Toggle Chrome's secure DNS setting, since a DNS-over-HTTPS provider can be unreachable on networks that block it.
  7. Check the hosts file for stale entries if one specific hostname fails while everything else resolves normally.

If you run the website

  1. Confirm your domain resolves from several external resolvers, not just from your own network where caches may be masking a problem.
  2. Check that every authoritative nameserver for the zone is answering, since one failing nameserver produces failures for a fraction of visitors.
  3. Verify the delegation at the registrar matches the nameservers you actually run, as a mismatch causes intermittent resolution failures.
  4. Keep an eye on domain expiry as well as DNS health, because a lapsed registration takes the whole zone down at once.
  5. Lower TTLs ahead of any planned DNS change so caches turn over quickly, and raise them again once the change has settled.

Still not fixed? Next steps

  • Test whether a raw IP loads while names fail. Reaching 1.1.1.1 in a browser while no hostname resolves isolates the fault to DNS immediately.
  • Set a public resolver explicitly on the adapter, then retry. Success there proves the previous configuration was the problem and not the network generally.
  • Flush the DNS cache and renew the DHCP lease, which clears stale entries and re-requests resolver addresses from the router.
  • Check for VPN leftovers in the adapter's DNS settings, particularly after a client crashed or was force-quit.
  • Inspect the hosts file for stale or unwanted entries when one specific hostname fails while everything else resolves.

Code & configuration

Copy-paste starting points. Replace example.com and the paths with your own, and test changes on staging before production.

Prove the resolver is the problem, not the name

# the configured resolver
dig example.com
# ;; connection timed out; no servers could be reached

# a public one, bypassing local configuration
dig @1.1.1.1 example.com +short
# 93.184.216.34   <- the name is fine

Clear cached state and re-request DNS settings

# Windows
ipconfig /flushdns
ipconfig /release && ipconfig /renew

# macOS
sudo dscacheutil -flushcache; sudo killall -HUP mDNSResponder

Check what resolvers the device is actually using

# macOS / Linux
scutil --dns | grep nameserver     # macOS
resolvectl status | grep "DNS Servers"

# Windows
ipconfig /all | findstr "DNS Servers"

How to prevent DNS_PROBE_FINISHED_BAD_CONFIG

The visitor-side version of this error never reaches you, but its counterpart does: a nameserver that stops answering or a zone that fails to load is invisible from inside your own network, because internal caches keep serving answers long after the authoritative source went quiet. Pulsetic checks your website from multiple locations every 30 seconds and alerts you by email, SMS, voice call, Slack, Discord, Telegram, or webhook the moment it stops resolving or responding from any of them, so a DNS failure is something you are told about rather than something you are informed of by a customer.

Learn how Pulsetic's domain monitoring detects this from the outside, across 15+ locations.

Sources and further reading

The specifications and vendor documentation this guide is written from, plus deeper reading on the parts it only summarises.

Frequently asked questions

  • What does DNS_PROBE_FINISHED_BAD_CONFIG mean?

    It means Chrome tried to resolve a domain name, failed, ran a diagnostic probe to find out why, and concluded that the DNS configuration on your device or network is not usable. No lookup could complete. It is not saying the website is missing or that the domain is wrong, only that the machinery responsible for turning names into addresses is not working where you are.

  • How do I fix DNS_PROBE_FINISHED_BAD_CONFIG?

    Set your DNS servers explicitly to a public resolver such as 1.1.1.1 or 8.8.8.8, flush the DNS cache, and retry. That resolves most cases. If a VPN disconnected recently, reset the network adapter, because VPN clients often leave DNS settings pointing at a resolver that is only reachable through the tunnel. If every device on the network is affected, restart the router.

  • How is this different from DNS_PROBE_FINISHED_NXDOMAIN?

    They are opposite verdicts from the same probe. NXDOMAIN means DNS worked correctly and gave a definitive answer: no such name exists, which usually means a typo or an expired domain. BAD_CONFIG means the lookup never completed at all, because the resolver settings are broken. NXDOMAIN points at the address you typed; BAD_CONFIG points at your network.

  • Is the website down if I see this error?

    Almost certainly not. The error describes a failure in your own DNS configuration, which happens before any contact with the website. The quickest way to confirm is to query a public resolver directly with dig @1.1.1.1 example.com. Getting an address back proves the domain is resolving perfectly well and that only your device could not find it.

  • Why does changing DNS servers fix it?

    Because it removes the broken component from the path. Devices normally take resolver addresses from the router over DHCP, and if those are wrong or point at something that has stopped responding, every lookup fails while the rest of the connection appears healthy. Setting a public resolver bypasses that entirely. It also doubles as a diagnosis: if the fix works, the fault was definitively in the DNS configuration.

  • Can a VPN cause this?

    Yes, and usually after it disconnects rather than while connected. VPN clients replace your DNS settings so lookups travel through the tunnel, and a client that crashes or is force-quit can leave those settings in place. You end up with a working connection whose resolver is only reachable inside a tunnel that no longer exists. Resetting the network adapter, or reconnecting and disconnecting cleanly, restores the original configuration.

  • Why do other apps work while Chrome does not?

    Because they may not be using the same resolver. Chrome can perform its own DNS-over-HTTPS lookups through its secure DNS setting, independently of the system configuration, and other applications often hold open connections or cached addresses that need no lookup at all. So a broken system resolver can stop the browser while a messaging app carries on. Toggling Chrome's secure DNS setting is worth trying when the browser alone is affected.

  • What is the difference between BAD_CONFIG and NXDOMAIN?

    They are opposite conclusions from the same probe. NXDOMAIN means DNS worked exactly as designed and returned a definitive answer: that name does not exist. BAD_CONFIG means the probe could not get a usable answer at all, because the resolver settings on the device or network are broken. One is a successful lookup with a negative result, the other is a lookup that never completed. That is why NXDOMAIN usually means a typo and BAD_CONFIG usually means your network.

  • Why does switching to 1.1.1.1 or 8.8.8.8 fix it?

    Because it bypasses whatever resolver was broken. Most devices take their DNS servers from the router over DHCP, and if those addresses are wrong, unreachable, or pointing at a resolver that has stopped responding, every lookup fails while the rest of the connection appears fine. Setting a public resolver explicitly removes that dependency. It is also a diagnosis rather than just a fix: if it works, you have proven the fault was in the DNS configuration rather than anywhere else.

  • Can a VPN cause DNS_PROBE_FINISHED_BAD_CONFIG?

    Frequently, and usually after it disconnects rather than while it is running. VPN clients replace your DNS settings so lookups go through the tunnel, and a client that exits badly can leave those settings behind, pointing at a resolver that is only reachable inside a tunnel that no longer exists. The symptom is a connection that looks perfectly healthy while no name resolves. Reconnecting and disconnecting the VPN cleanly, or resetting the adapter, restores the original settings.

  • Why do other apps work while the browser does not?

    Usually because they are not all using the same resolver. Chrome can use DNS-over-HTTPS through its own secure DNS setting, which is independent of the system configuration, and some applications keep their own caches or connect to addresses directly without a lookup. So a broken system resolver can break the browser while a chat app with an open connection carries on. Toggling Chrome's secure DNS setting is worth trying for exactly this reason.

  • Does the hosts file matter here?

    It can. The hosts file is consulted before DNS, so a corrupted or leftover entry can point a hostname at an address that no longer works, or interfere with resolution more generally. Entries left behind by development work, ad-blocking lists, or software that was uninstalled untidily are the usual sources. It is worth checking when a single specific website fails while everything else resolves normally.

Trusted by teams at companies around the world