(Podman) Containers in pod can't find each other anymore?

Rocky Linux General
1

Hey,

I’ve been running Rockylinux on a small server for about 2 months, and I used to have a bunch of rootless containers working flawlessly. However, recently, everything has started to fall apart. I have automatic updates turned on, so idk if the problem may be due to a recent update.

As an example, I was running Bookstack using images from linuxserver.io. This requires a database and the app itself, so I had them both in a pod so they could communicate.

Here’s the script I use to create the pod and containers:

DB_USER=bstack
PASS="randomcrap"
ROOTPASS="randomcrap"

DBNAME=bookstackapp

podman pod create --name bookstack -p 6875:80

#Database
podman run \
        --detach \
        --name bookstack_db \
        --pod bookstack \
        -e PUID=1000 \
        -e PGID=1000 \
        -e TZ=America/New_York \
        -e MYSQL_DATABASE=$DBNAME \
        -e MYSQL_USER=$DB_USER \
        -e MYSQL_PASSWORD=$PASS \
        -e MYSQL_ROOT_PASSWORD=$ROOTPASS \
        -v /myserver/bookstack/database:/config:Z \
        --restart unless-stopped \
        lscr.io/linuxserver/mariadb

#Bookstack
podman run \
        --detach \
        --name bookstack_app \
        --pod bookstack \
        -e PUID=1000 \
        -e PGID=1000 \
        -e APP_URL=http://myserver:6875 \
        -e DB_HOST=bookstack_db \
        -e DB_USER=$DB_USER \
        -e DB_PASS=$PASS \
        -e DB_DATABASE=$DBNAME \
        -v /myserver/bookstack/app:/config:Z \
        --restart unless-stopped \
        lscr.io/linuxserver/bookstack

And that was working flawlessly for a few weeks, but now it’s failing because the bookstack_app container can’t find bookstack_db. The logs display the following error repeatedly:

nc: getaddrinfo: Name does not resolve

nslookup fails with the following:

$ podman exec bookstack_app nslookup bookstack_db
Server: 10.0.2.3
Address: 10.0.2.3:53

** server can’t find bookstack_db.local: NXDOMAIN

** server can’t find bookstack_db.local: NXDOMAIN

Besides that, networking works fine, and containers can contact the outside world without issue.

Again, this was all working flawlessly just a few days ago. I’ve tried running podman system reset, rebooting, and recreating everything, but the issue persists. It affects all my pods (not just the bookstack example above). There are also no SELinux errors.

Anyone know what’s up?

Thanks!

2

There is no bookstack_db (nor bookstack_app ) network namespace.
Containers inside a pod should communicate via localhost, from outside a pod via bookstack (pod name).

3

Sorry for the late response, but thanks for the help.

It seems like switching to localhost fixes the connectivity issue I was observing, but it’s weird because it was working before using those hostnames. Those containers were up for weeks with no issues, and I even restarted them a few times due to power outages.

The bookstack app appears to be having some other issue (possibly related? I’ll assume not tho), but a Quay instance I had on another pod did start working again. :thinking: