To run Linkwarden non-root with Docker, create the data , pgdata and meili_data folders yourself before the first docker compose up , then add user: "1000:1000" to all three services ( linkwarden , postgres , meilisearch ) in docker-compose.yml . For an existing install, stop the stack and run sudo chown -R 1000:1000 data pgdata meili_data first. That is the method in the official Running as a Non-Root User guide added with Linkwarden v2.16.1 . Below is the whole setup with real output, plus the two mistakes that leave the stack in a restart loop. Tested on 2026-10-27 on a Debian 12 KVM virtual machine (4 vCPU, 4 GB RAM) with Docker Engine 29.9.2 and Docker Compose v5.6.1, using ghcr.io/linkwarden/linkwarden:v2.16.3 (newest release that day), postgres:16-alpine and getmeili/meilisearch:v1.13.3 from the upstream docker-compose.yml . The host user is vagrant with UID and GID 1000. Every output below is copied from that lab; links were added through the Linkwarden API...

A docker mailserver upgrade from 15.x to 16.x is not just a new image tag. Version 16.0.0 moves the image from Debian 12 to Debian 13 and from Dovecot 2.3 to Dovecot 2.4. Your mail and accounts survive the switch, but an old dovecot.cf with 2.3 setting names stops Dovecot completely, SA_SPAM_SUBJECT is rejected, and DHE ciphers disappear from ports 587 and 993. The container still reports "up and running". Below is the full upgrade on a test server, what broke and the exact fixes. Lab: docker-mailserver 15.1.0 upgraded to 16.0.1 on the same volumes, rootless podman 5.8 (the docker commands are identical), a Debian 13 client container with swaks, curl and openssl, October 2026. Every output below comes from that lab. Sources: the docker-mailserver v16.0.0 release , its CHANGELOG and the Dovecot 2.3 to 2.4 upgrade guide . Docker mailserver upgrade to v16, in short: Back up docker-data/dms/ and stop the container. Rename 2.3 settings in config/dovecot.cf (for exampl...

Since docker-mailserver v16.0.0 the image ships a docker mailserver healthcheck out of the box: a script called dms-healthcheck and a HEALTHCHECK --start-period=30s CMD dms-healthcheck line in the Dockerfile. The script checks that every supervisor service your configuration enables (Postfix, Dovecot, cron, rsyslog, OpenDKIM, Rspamd and so on) is RUNNING . On Docker the container then shows (healthy) or (unhealthy) with no extra configuration. On podman the image healthcheck is not used at all, so you have to add --health-cmd dms-healthcheck yourself. Every example below was tested on 2026-10-22 with mailserver/docker-mailserver:16.0.1 and :15.1.0 on Docker Engine 29.9.1 (the official docker:29.9.1-dind image) with Docker Compose v5.6.0, and with ghcr.io/docker-mailserver/docker-mailserver:16.0.1 on rootless podman 5.8.9. Rspamd, OpenDKIM, OpenDMARC, Amavis, ClamAV, Fail2Ban and Postgrey were disabled to keep the lab small. All output is copied from the lab. Contents What ...

The quickest WireGuard Docker setup is the lscr.io/linuxserver/wireguard image with a short compose.yaml : set PEERS and SERVERURL , publish UDP port 51820, add the NET_ADMIN capability and the src_valid_mark sysctl, then run docker compose up -d . The container generates the server config and one ready-to-import client config (plus QR code) per peer. This guide shows the whole setup with real output, a working client, and the four mistakes that leave you with a tunnel that never connects. Tested on 2026-10-23 on a Debian 12 KVM virtual machine (kernel 6.1.0-47-amd64) with Docker Engine 29.9.2 and Docker Compose v5.6.1, using image tag 1.0.20260223-r0-ls127 , the newest release that day. A second container on the same VM played the role of a remote laptop and connected to the VM's IP 192.168.121.137. Private and preshared keys are shortened in the output below; everything else is copied from the lab. Contents What you need for WireGuard Docker Step 1: compose.yaml for the W...

Docker Engine 29.8.0 added an annotation filter to the container list. Run docker ps --filter annotation=KEY to list containers that carry an annotation key, or docker ps --filter annotation=KEY=VALUE to match the key and the exact value. Older engines (29.7.x and before) store annotations set with docker run --annotation , but answer the same filter with Error response from daemon: invalid filter 'annotation' . Every example below was tested on 2026-10-20 with the official images docker:29.9.1-dind (API 1.56) and docker:29.7.2-dind (API 1.55), each running in a rootless podman container. All output is copied from that lab; nothing is shortened except long container IDs. Contents What changed in Docker 29.8.0 Annotations vs labels docker ps --filter annotation: key and key=value Exact match only: no wildcards, case-sensitive Empty values Combining docker ps --filter annotation with other filters The label filter does not see annotations Use the filter with stop, inspec...

In the vast, ever-evolving landscape of modern software development, few technologies have made as profound an impact as containerization. And at the heart of the container revolution lies Kubernetes, the open-source system that has become synonymous with container orchestration. What began as an internal project at Google, dubbed Borg, evolved into a phenomenon that now powers everything from small startups to Fortune 500 enterprises. This isn't just a tool; it's a paradigm shift, fundamentally changing how applications are built, deployed, and managed in an increasingly distributed world. This deep dive aims to demystify Kubernetes, taking you from its foundational concepts to advanced considerations, helping you understand not just what it is, but why it became indispensable. Whether you're a seasoned DevOps engineer, a curious developer, or an IT leader planning your next infrastructure move, prepare to immerse yourself in the world of resilient, scalable, and highly...