#Backup
Secrets Set to 600, but 0644 Logs Go Offsite Every Night
The .env is encrypted and permissions are locked down on every secret file, but customer data still leaks off the machine every night through something nobody classifies as confidential: logs that every account on the machine can read, bundled into backups uploaded to object storage. Here is how I stop it at the source by logging only result codes.
Every File Is Backed Up, but I Still Can't Rebuild the Server
The database dump is complete, the code tar is complete, and all the secrets are there, but put them on a blank machine and the site will not come up, because the things that actually make the server work were never in the backup files. This article lists what disappears with the machine and the recovery sequence I wrote for my future self.
Backups Succeeded Nightly, but Offsite Uploads Failed for 3 Weeks
cron ran every night, local files were complete every day, and the log said done, but offsite storage had not changed for 3 weeks. The first 8 nights of backups were already gone for good because local rotation had deleted them first. This article examines the two gaps that let the failure stay silent for so long.
Recover MariaDB Redo Log Corruption with force_recovery and Reinit
MariaDB would not start after an unclean shutdown — a corrupted InnoDB redo log took the mysql.global_priv system table down with it. This is the recovery sequence I actually used, starting with innodb_force_recovery to make the data readable, then dumping each db before reinitializing the whole system, with one rule never to raise force_recovery higher than necessary.
Set DirectAdmin Admin Backup Cron from the CLI Without the Panel
CMD_ADMIN_BACKUP takes five steps to set up a cron job, but DirectAdmin actually just appends one line to backup_crons.list and lets dataskq pick it up. Here is how to write directly to the file from the CLI, including common traps with when=now, base64, and append_to_path.
Recovering Old Website Images with Wayback Machine CDX API + Python
A website over a decade old had broken images throughout because it hotlinked to a dead CDN — I wrote a Python script to list images from the Wayback Machine CDX API and restore them to their original paths without touching a single character of HTML, with the full script I actually used.
Managing Linux Servers Safely with AI over SSH
How I actually use Claude to audit and fix Linux server security over SSH, from preparing my machine and storing secrets to hard rules before touching production
Off-Site Backups with Pull Architecture + rclone/rsync
Add off-site Layer 3 to complete 3-2-1 with pull architecture (the destination host fetches the data), rclone + rsync, and lessons from real pitfalls involving NAT, RSA host keys, cwRsync, and PID recycling.
Rebuilding DirectAdmin Backups with NFS and POSIX ACL in Production
A record of rebuilding backups on a DirectAdmin server after migrating to a new machine — from FTP push that broke at three layers to an NFSv4 hard mount + POSIX ACL, with kernel-level lessons learned along the way.
Moving DA from CentOS 7 to Debian 12, Part 4 — Cutover and 17 Issues
The actual cutover, where the near-0 downtime plan fell apart because of the DA Legacy license and slow qcow2 reads, ended with 1 hour of downtime and 17 issues in quick succession after the switch — critical web stack outages, medium issues affecting subsets of services, and minor debugging notes.
Moving DA to Debian 12, Part 2 — Switching Storage to a ZFS Mirror
Replace the three storage layers of mdadm RAID1 + ext4 + qcow2 with a single layer of ZFS mirror + raw zvol while the old VM stays online, by degrading RAID to a single disk.