Debug Prometheus target relabeling before samples disappear
Prometheus relabel_configs: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
PromQL rate vs irate: choose by alerting and graph resolution
PromQL rate vs irate: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Resize an ext4 filesystem after verifying the device boundary
Resize ext4 filesystem: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Grow an XFS filesystem from its mount point, not a guessed device
Resize XFS filesystem: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
rsync --partial vs --inplace: choose by recovery and reader behavior
Rsync partial vs inplace: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use SFTP batch mode with explicit failure and atomic-file behavior
SFTP batch mode: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Build SLO burn-rate alerts from good and total events
SLO burn rate alerts: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use ss to identify the process listening on a Linux port
Ss command listening ports: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use SSH BatchMode so automation fails instead of waiting for input
SSH BatchMode: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use SSH Match exec without turning config into an opaque program
SSH Match exec: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use SSH ControlPersist in automation without stale shared masters
SSH ControlPersist automation: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use ExitOnForwardFailure so SSH tunnels fail loudly
ExitOnForwardFailure: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use ssh-keyscan for collection, never as automatic verification
Ssh-keyscan safe use: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Automate SSH ProxyJump without placing private keys on the bastion
SSH ProxyJump automation: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
ServerAliveInterval vs TCPKeepAlive for SSH connection liveness
ServerAliveInterval vs TCPKeepAlive: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use StrictHostKeyChecking accept-new with a clear trust model
StrictHostKeyChecking accept-new: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
systemctl daemon-reload vs restart: what each command changes
Systemctl daemon-reload vs restart: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Pass secrets with systemd credentials instead of environment variables
Systemd credentials secrets: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Set systemd service environment variables without hidden drift
Systemd environment variables: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use network-online.target without hiding application retries
Systemd network-online.target: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Debug systemd-resolved by link, search domain, and server
Systemd-resolved troubleshooting: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Set systemd CPU and memory limits with observable headroom
Systemd resource limits: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Choose a systemd Restart policy without creating a crash loop
Systemd Restart policy: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.
Use systemd socket activation without two processes fighting for a port
Systemd socket activation: inspect the effective state, interpret expected and abnormal results, make one bounded change, verify, and roll back safely.