Record full error
Get the first real error message with time information in the journal, service and kernel log instead of the general warning on the screen.
For DNS troubleshooting, first check whether the server can reach the outside world at the IP layer. If access to the IP works but the domain does not resolve, it is the resolver; if IP access is also unavailable, there is a route, gateway, firewall or interface problem.
ping: example.com: Temporary failure in name resolution
curl: (6) Could not resolve host: example.com
Network is unreachable
For DNS troubleshooting, first check whether the server can reach the outside world at the IP layer. If access to the IP works but the domain does not resolve, it is the resolver; if IP access is also unavailable, there is a route, gateway, firewall or interface problem.
Get the first real error message with time information in the journal, service and kernel log instead of the general warning on the screen.
Test service, disk, network, permission, security policy and application configuration separately without changing them at the same time.
The service name, package manager and firewall tool may be different on Ubuntu/Debian and AlmaLinux/Rocky/CloudLinux.
Keep a second session, console, snapshot or current backup for SSH, network, firewall, boot and disk changes.
Do not modify the /etc/resolv.conf file without knowing the symlink structure used by the administrator service.
Meaning: The resolver cannot convert the domain name to an IP.
Possible cause: DNS server, resolv.conf or network problem.
Meaning: The application did not receive a DNS response.
Possible cause: Wrong nameserver, proxy or container DNS.
Meaning: Hedefe rota yok.
Possible cause: Default route or interface is down.
Meaning: The gateway or local network cannot reach the target.
Possible cause: Wrong subnet/gateway or link problem.
Meaning: The local stub resolver is not working.
Possible cause: Service or symlink is broken.
Meaning: The manually written DNS was changed by DHCP or an admin tool.
Possible cause: Editing outside Netplan/NetworkManager.
Meaning: The authoritative or recursive DNS could not complete the query.
Possible cause: DNSSEC, upstream or zone issue.
Meaning: The local hostname does not match /etc/hosts.
Possible cause: The hostname changed but hosts was not updated.
No error records were found matching this expression.
ip -br addr
ip route
ip rule
Shows IP addresses, default route and policy routes.
ping -c 3 1.1.1.1
getent ahosts example.com
dig +time=2 +tries=1 example.com
Separates network access from DNS resolution.
resolvectl status 2>/dev/null || true
readlink -f /etc/resolv.conf
cat /etc/resolv.conf
Shows the systemd-resolved and resolv.conf structure.
netplan get 2>/dev/null || true
Shows the persistent network and nameserver configuration.
nmcli device status 2>/dev/null || true
nmcli device show 2>/dev/null | grep -E 'GENERAL.DEVICE|IP4.DNS|IP4.GATEWAY|IP4.ADDRESS'
Shows the active DNS and gateway values on the RHEL family.
nc -zvu -w2 1.1.1.1 53
nc -zv -w2 1.1.1.1 53
Tests UDP and TCP port 53 access.
The Netplan, systemd-networkd or NetworkManager backend must be determined; the /etc/resolv.conf symlink structure must be preserved.
networkctl status 2>/dev/null || truesystemctl status systemd-resolved --no-pager 2>/dev/null || truenetplan get 2>/dev/null || trueThe ipv4.dns, ignore-auto-dns and gateway settings inside the NetworkManager connection profile should be used.
nmcli connection show --activenmcli device showsystemctl status NetworkManager --no-pagerDocker, container resolver, panel DNS service and host resolver must be separated from each other.
docker info 2>/dev/null | grep -i dns || truecat /etc/docker/daemon.json 2>/dev/null || truesystemctl status named bind9 pdns --no-pager 2>/dev/null || trueRecord the exact on-screen or logged error message with its timestamp, then test the service, network, disk, and security layers separately.
The package manager, firewall, network manager, and some service names differ by distribution family. The distribution section on the page explains these differences.
A restart may restore service temporarily, but restarting before collecting logs and resource data can hide the root cause.
Read-only diagnostic commands are generally safe. Configuration, firewall, disk, and recovery changes require console access, a backup, and a rollback plan.
The Linux layer is the same, but control-panel service wrappers, firewall plugins, and vendor packages must also be considered.
Keep a second session and the provider console open during SSH, firewall, network, or boot changes.
No. Technical accuracy and coverage provide a strong foundation, but rankings also depend on competition, site authority, internal links, speed, and user behavior.
We analyze Ubuntu, Debian, AlmaLinux, Rocky and CloudLinux service, disk, network, security, performance and boot problems with log data.