How to setup volume for postgres log folder - Permission denied error - postgresql

Currently, I would like to mount logging folder for postgres, so that even after host machine restart, logging information is still preserved.
I use the following docker-compose.yml
docker-compose.yml
version: '2'
services:
postgres:
image: postgres:latest
restart: always
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=55F4rGFwsXXXXXX
ports:
- "5432:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
- postgres_logs:/logs
# Make Postgres log to a file.
# More on logging with Postgres: https://www.postgresql.org/docs/current/static/runtime-config-logging.html
command: postgres -c logging_collector=on -c log_destination=stderr -c log_directory=/logs
volumes:
postgres_data:
postgres_logs:
However, I'm getting the following permission denied error.
postgres_1 | 2018-02-19 09:03:45.359 UTC [1] LOG: database system is shut down
postgres_1 | 2018-02-19 09:04:45.963 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres_1 | 2018-02-19 09:04:45.963 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres_1 | 2018-02-19 09:04:45.965 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres_1 | 2018-02-19 09:04:45.972 UTC [1] FATAL: could not open log file "/logs/postgresql-2018-02-19_090445.log": Permission denied
postgres_1 | 2018-02-19 09:04:45.974 UTC [1] LOG: database system is shut down
postgres_1 | 2018-02-19 09:05:46.741 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres_1 | 2018-02-19 09:05:46.741 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres_1 | 2018-02-19 09:05:46.744 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres_1 | 2018-02-19 09:05:46.753 UTC [1] FATAL: could not open log file "/logs/postgresql-2018-02-19_090546.log": Permission denied
postgres_1 | 2018-02-19 09:05:46.755 UTC [1] LOG: database system is shut down
postgres_1 | 2018-02-19 09:06:47.366 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres_1 | 2018-02-19 09:06:47.366 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres_1 | 2018-02-19 09:06:47.368 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres_1 | 2018-02-19 09:06:47.375 UTC [1] FATAL: could not open log file "/logs/postgresql-2018-02-19_090647.log": Permission denied
postgres_1 | 2018-02-19 09:06:47.377 UTC [1] LOG: database system is shut down
Anyone has idea how can I resolve such error?

2nd Edit
After some reading it seems that is not possible to do this the way you need. You would need to be able to define file ownership when declaring a volume with docker-compose, and this is something that is not supported by the docker engine. But there are a few workarounds that you can consider, check here for more details.
As a workaround you could create a Dockerfile extending postgres and add this:
# ...
RUN mkdir /logs
RUN chown postgres:postgres /logs
Edit
After some experimentation this is working:
version: '2'
services:
postgres:
image: postgres:latest
restart: always
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=55F4rGFwsXXXXXX
ports:
- "5432:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
- ./:/logs:z
# Make Postgres log to a file.
# More on logging with Postgres: https://www.postgresql.org/docs/current/static/runtime-config-logging.html
command: postgres -c logging_collector=on -c log_destination=stderr -c log_directory=/logs
volumes:
postgres_data:
Original answer
I accidentally got it working by doing this, first running docker-compose up with this docker-compose file:
version: '2'
services:
postgres:
image: postgres:latest
restart: always
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=55F4rGFwsXXXXXX
ports:
- "5432:5432"
volumes:
- postgres_data:/var/lib/postgresql/data
- postgres_logs:/logs:z
# Make Postgres log to a file.
# More on logging with Postgres: https://www.postgresql.org/docs/current/static/runtime-config-logging.html
command: /bin/bash -c "mkdir -p /logs && chmod -R 777 /logs && postgres -c logging_collector=on -c log_destination=stderr -c log_directory=/logs"
volumes:
postgres_data:
postgres_logs:
This fails with:
postgres_1 | "root" execution of the PostgreSQL server is not
permitted. postgres_1 | The server must be started under an
unprivileged user ID to prevent postgres_1 | possible system security
compromise. See the documentation for postgres_1 | more information
on how to properly start the server.
After reverting the changes to command and running again then works, but obviously this is not a solution so stick with the edited answer above.

the problem is the permission in host, not in container.
so in host,
chmod -R 777 postgres_logs

Related

docker compose psql: error: FATAL: role "postgres" does not exist

I faced a problem when I try to use psql command with my docker-compose file on my local Ubuntu machine:
psql: error: FATAL: role "postgres" does not exist
I tried to use others solution like removing docker image, volume. psql -U postgres doesn't work for me either.
I try to use first docker-compose up, then docker exec -it database bash
There's my docker-compose file
services:
db:
container_name: postgres
image: postgres:13.3-alpine
restart: always
user: postgres
environment:
- POSTGRES_DB=postgres
- POSTGRES_PASSWORD=postgres
- POSTGRES_USER=root
ports:
- "5432:5432"
volumes:
- ./data/db:/var/lib/postgresql/data
Maybe this string tells something?
postgres | PostgreSQL Database directory appears to contain a database; Skipping initialization
OUTPUT:
Attaching to postgres
postgres |
postgres | PostgreSQL Database directory appears to contain a database; Skipping initialization
postgres |
postgres | 2021-08-02 17:29:10.426 UTC [1] LOG: starting PostgreSQL 13.3 on x86_64-pc-linux-musl, compiled by gcc (Alpine 10.3.1_git20210424) 10.3.1 20210424, 64-bit
postgres | 2021-08-02 17:29:10.426 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres | 2021-08-02 17:29:10.426 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres | 2021-08-02 17:29:10.429 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres | 2021-08-02 17:29:10.433 UTC [12] LOG: database system was shut down at 2021-08-02 17:22:17 UTC
postgres | 2021-08-02 17:29:10.438 UTC [1] LOG: database system is ready to accept connections
postgres | 2021-08-02 17:37:53.452 UTC [33] FATAL: role "postgres" does not exist
postgres | 2021-08-02 17:37:56.958 UTC [35] FATAL: role "user" does not exist
postgres | 2021-08-02 17:41:54.294 UTC [45] FATAL: role "postgres" does not exist```
First, you've set POSTGRES_USER to root, so you're going to have a root user instead of postgres user.
Second, if a database already exists, it doesn't matter what you set for POSTGRES_USER and POSTGRES_PASSWORD -- postgres will use whatever is in the database.
So you can either:
Delete the database (rm -rf data/db) and start over, or
Edit your pg_hba.conf so that you don't need a password
docker-compose exec web rake db:migrate db:create - I have above error: role "..." does not exist.
When I try:
docker-compose exec web rake db:drop
docker-compose exec web rake db:setup
it's working.

Postgres database connectivity issues when using docker-compose

I am trying to dockerize the REST api written in golang, but having troubles connecting to postgres database (the app works fine without docker). Below are my Dockerfile and docker-compose.yml
FROM golang:alpine
RUN mkdir /app
ADD . /app
WORKDIR /app
COPY go.mod .
RUN go mod download
COPY . .
RUN go build -o main .
CMD ["/app/main"]
docker-compose.yml
version: '3'
volumes:
postgres_data:
driver: local
services:
postgres:
image: postgres
environment:
POSTGRES_DB: test_db
POSTGRES_USER: postgres
POSTGRES_PASSWORD: 1234
ports:
- 5432:5432
my-go-app:
restart: always
build:
dockerfile: Dockerfile
context: .
environment:
# Postgres Details
DB_HOST: postgres
DB_DRIVER: postgres
DB_USER: postgres
DB_PASSWORD: 1234
DB_NAME: test_db
DB_PORT: 5432
# APP details
APP_PROTOCOL: http
APP_HOST: localhost
APP_PORT: 10000
ALLOWED_ORIGINS: "*"
ports:
- 10000:10000
depends_on:
- postgres
Database properties: host = "localhost", port = 5432, user = "postgres", password="1234", dbname="test_db" (all specified in docker-compose.yml). App itself runs on port 10000
When running docker up and docker-compose up --build I'm getting sql: database is closed error
Starting rest_db_my-go-app_1 ... done
Attaching to rest_db_postgres_1, rest_db_my-go-app_1
my-go-app_1 | 2021/04/02 12:54:07 Connection Failed to Open
my-go-app_1 | 2021/04/02 12:54:07 Starting development server at http://127.0.0.1:10000/
my-go-app_1 | 2021/04/02 12:54:07 Quit the server with CONTROL-C.
my-go-app_1 |
my-go-app_1 | (/app/main.go:185)
my-go-app_1 | [2021-04-02 12:54:07] sql: database is closed
postgres_1 |
postgres_1 | PostgreSQL Database directory appears to contain a database; Skipping initialization
postgres_1 |
postgres_1 | 2021-04-02 12:02:15.807 UTC [1] LOG: starting PostgreSQL 13.2 (Debian 13.2-1.pgdg100+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 8.3.0-6) 8.3.0, 64-bit
postgres_1 | 2021-04-02 12:02:15.808 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres_1 | 2021-04-02 12:02:15.808 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres_1 | 2021-04-02 12:02:15.815 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres_1 | 2021-04-02 12:02:15.831 UTC [26] LOG: database system was shut down at 2021-04-02 12:02:14 UTC
postgres_1 | 2021-04-02 12:02:15.837 UTC [1] LOG: database system is ready to accept connections
postgres_1 | 2021-04-02 12:02:25.830 UTC [1] LOG: received fast shutdown request
postgres_1 | 2021-04-02 12:02:25.834 UTC [1] LOG: aborting any active transactions
postgres_1 | 2021-04-02 12:02:25.835 UTC [1] LOG: background worker "logical replication launcher" (PID 32) exited with exit code 1
postgres_1 | 2021-04-02 12:02:25.835 UTC [27] LOG: shutting down
postgres_1 | 2021-04-02 12:02:25.862 UTC [1] LOG: database system is shut down
How to solve this database connectivity issue? Thanks in advance

Docker: Adminer backend for PostgrSQL database times out

I'm trying to set up a simple database system using a PostgreSQL Docker container and an Adminer container as a backend. Note this is all running on a RaspberryPi 4.
docker-compose.yml:
services:
postgres:
image: postgres
restart: on-failure
ports:
- 5432:5432
env_file:
- ./.auth_file
volumes:
- db:/var/lib/postgresql/data
adminer:
image: adminer
restart: on-failure
ports:
- 8070:8080
env_file:
- ./.auth_file
depends_on:
- postgres
volumes:
db:
I then run it using docker-compose up:
postgres_1 |
postgres_1 | PostgreSQL Database directory appears to contain a database; Skipping initialization
postgres_1 |
postgres_1 | 2021-04-01 13:25:25.919 UTC [1] LOG: starting PostgreSQL 13.2 (Debian 13.2-1.pgdg100+1) on arm-unknown-linux-gnueabihf, compiled by gcc (Debian 8.3.0-6) 8.3.0, 32-bit
postgres_1 | 2021-04-01 13:25:25.919 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
postgres_1 | 2021-04-01 13:25:25.919 UTC [1] LOG: listening on IPv6 address "::", port 5432
postgres_1 | 2021-04-01 13:25:26.028 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
postgres_1 | 2021-04-01 13:25:26.090 UTC [27] LOG: database system was shut down at 2021-04-01 13:20:41 UTC
postgres_1 | 2021-04-01 13:25:26.116 UTC [1] LOG: database system is ready to accept connections
adminer_1 | [Sun Jun 14 00:31:20 2071] PHP 7.4.16 Development Server (http://[::]:8080) started
adminer_1 | [Sun Jun 14 00:30:16 2071] [::ffff:192.168.16.1]:50886 Accepted
The last line shows that I tried to access localhost:8070 in the Browser right on the Pi. The page just loads forever and Adminer never returns the page.
Could this have to do with some firewall issues?
Thanks for any help!
Found the solution. This has to do with the Docker Container being unable to fetch the current time from the host, due to some version issues surrounding Docker, Alpine (the base image for Adminer) and libseccomp. This is also evident by the fact that Adminer shows entirely different timestamps than the Postgres-Container in my original post.
The full explanation can be found here: https://wiki.alpinelinux.org/wiki/Release_Notes_for_Alpine_3.13.0#time64_requirements
After following the steps linked above, the docker-compose.yml should look something like this:
adminer:
image: adminer
...
security_opt:
- seccomp=config/compose/default.json # required to avoid clock_gettime() errors
Replace config/compose/default.json with the relative path to the modified default.json.

Can't connect Go Application to the postgres container in docker-compose

I have been trying to connect my go application to postgresql db using docker-compose.
Here is my Dockerfile for the web server
FROM golang:latest
RUN mkdir /app
COPY . /app
WORKDIR /app
RUN go mod download
EXPOSE 5000
CMD ["go","run","main.go"]
And this is my docker-compose.yml file
version: '3'
services:
db:
image: postgres:latest
restart: always
network_mode: bridge
environment:
POSTGRES_USER: factly
POSTGRES_PASSWORD: qwertyui
POSTGRES_DB: factly
ports:
- "5432:5432"
server:
build: .
depends_on:
- db
ports:
- "5000:5000"
But I can't connect to the postgresql db as it gives a connection refused error message from the
Go serverStarting factly_db_1 ...
Starting factly_db_1 ... done
Recreating factly_server_1 ...
Recreating factly_server_1 ... done
Attaching to factly_db_1, factly_server_1
db_1 |
db_1 | PostgreSQL Database directory appears to contain a database; Skipping initialization
db_1 |
db_1 | 2020-08-19 06:11:17.151 UTC [1] LOG: starting PostgreSQL 12.4 (Debian 12.4-1.pgdg100+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 8.3.0-6) 8.3.0, 64-bit
db_1 | 2020-08-19 06:11:17.151 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
db_1 | 2020-08-19 06:11:17.152 UTC [1] LOG: listening on IPv6 address "::", port 5432
db_1 | 2020-08-19 06:11:17.266 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
db_1 | 2020-08-19 06:11:17.906 UTC [26] LOG: database system was shut down at 2020-08-19 05:56:39 UTC
db_1 | 2020-08-19 06:11:18.213 UTC [1] LOG: database system is ready to accept connections
server_1 | host=127.0.0.1 port=5432 user=factly password=qwertyui dbname=factly sslmode=disable
server_1 | 2020/08/19 06:11:21 dial tcp 127.0.0.1:5432: connect: connection refused
server_1 | exit status 1
factly_server_1 exited with code 1
I have tried running just the postgres container and connecting it with the Go Application which works perfectly fine but running the Go Application inside a container gives the same error(Which makes me feel the Go web server container couldn't connect outside of it's network).
How can I correct this?
I guess the problem is you are trying to connect to 127.0.0.1 from the go program instead use the hostname of the database container db.

How to use Netcat to check whether postgresql docker container is up

I want to see if docker container with PostgreSQL is ready using Netcat utility.
My entrypoint.sh script seems to be unable to spot the DB up and running.
When I login into docker and run NC in verbose mode I get
DNS fwd/rev mismatch: db != telegram_messages_db_1.telegram_messages_default
What am I doing wrong?
My setup
entrypoint.sh
echo "Waiting for postgres..."
while ! nc -z db 5432; do
sleep 0.1
done
echo "PostgreSQL started"
docker-compose.yaml
version: '3.7'
services:
messages:
build:
context: .
dockerfile: Dockerfile
entrypoint: ['/usr/src/app/entrypoint.sh'] # new
volumes:
- .:/usr/src/app
ports:
- 5001:5000
environment:
.....
depends_on:
- db
db:
build:
context: ./backend/db
dockerfile: Dockerfile
expose:
- 5432
environment:
- POSTGRES_USER=postgres
- POSTGRES_PASSWORD=postgres
docker-compose output
Attaching to telegram_messages_db_1, telegram_messages_messages_1
messages_1 | Waiting for postgres...
db_1 |
db_1 | PostgreSQL Database directory appears to contain a database; Skipping initialization
db_1 |
db_1 | 2020-03-29 23:55:22.103 UTC [1] LOG: starting PostgreSQL 12.2 on x86_64-pc-linux-musl, compiled by gcc (Alpine 9.2.0) 9.2.0, 64-bit
db_1 | 2020-03-29 23:55:22.103 UTC [1] LOG: listening on IPv4 address "0.0.0.0", port 5432
db_1 | 2020-03-29 23:55:22.104 UTC [1] LOG: listening on IPv6 address "::", port 5432
db_1 | 2020-03-29 23:55:22.115 UTC [1] LOG: listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
db_1 | 2020-03-29 23:55:22.170 UTC [20] LOG: database system was shut down at 2020-03-29 23:46:56 UTC
db_1 | 2020-03-29 23:55:22.190 UTC [1] LOG: database system is ready to accept connections
Also...
If I run nc -z -v telegram_messages_db_1.telegram_messages_default 5432
I do get a nice response:
telegram_messages_db_1.telegram_messages_default [172.19.0.2] 5432 (postgresql) open
Removing all containers, rebooting and building from scratch solved the issue.
I will leave it here for posterity.