I've done a fresh install of Supabase via docker compose following this example into my Ubuntu 20.04 server. When I run docker-compose up everything works except I see this error in my logs:
supabase-rest | 21/Feb/2022:19:50:29 +0000: {"details":"could not translate host name \"l#db\" to address: Name does not resolve\n","code":"","message":"Database connection error. Retrying the connection."}
How do I fix this?
Looking at this section in the docker-compose.yml:
rest:
container_name: supabase-rest
image: postgrest/postgrest:v9.0.0
depends_on:
- db
restart: unless-stopped
environment:
PGRST_DB_URI: postgres://postgres:${POSTGRES_PASSWORD}#db:5432/postgres
It's possible that the password you used in .env is not being escaped correctly - does it contain special chars?
Note that they need to be URI encoded if so as per: How to handle special characters in the password of a Postgresql URL connection string?
Related
I have local server with domain mydomain.com it is just alias to localhost:80
And I want to allow make requests to mydomain.com from my running docker-container.
When I'm trying to request to it I see
cURL error 7: Failed to connect to mydomain.com port 80: Connection refused
My docker-compose.yml
version: '3.8'
services:
nginx:
container_name: project-nginx
image: nginx:1.23.1-alpine
volumes:
- ./docker/nginx/conf.d/default.conf:/etc/nginx/conf.d/default.conf
- ./src:/app
ports:
- ${NGINX_PORT:-81}:80
depends_on:
- project
server:
container_name: project
build:
context: ./
environment:
NODE_MODE: service
APP_ENV: local
APP_DEBUG: 1
ALLOWED_ORIGINS: ${ALLOWED_ORIGINS:-null}
volumes:
- ./src:/app
I'm using docker desktop for Windows
What can I do?
I've tried to add
network_mode: "host"
but it ruins my docker-compose startup
When I'm trying to send request to host.docker.internal I see this:
The requested URL was not found on this server. If you entered
the URL manually please check your spelling and try again.
The host network is not supported on Windows. If you are using Linux containers on Windows, make sure you have switched to Linux containers on Docker Desktop. That uses WSL2, so you should be able to use that in there.
I have set an Oracle docker image (https://github.com/oracle/docker-images/tree/main/OracleDatabase/SingleInstance/dockerfiles) which by default is running on port 1521.
I would like to change the port in the Image to 1531.
I know that in the docker-compose I can set "1531:1521" BUT the other container still searching for port 1521 in the created network.
I tried to modify the port referenced in the Dockerfile of the version I want to use (19.3.0) and also in the createDB.sh but when I try to connect with the SID it fails, the listener is not working as expected.
Anybody already succeeded?
Update 1:
Here is the error message when I try to connect to the database after I changed the port.
SQL> CONNECT sys/HyperSecuredPassword#ORCLCDB AS sysdba; ERROR: ORA-12514: TNS:listener does not currently know of service requested in connect descriptor
Update 2:
I have the following docker-compose.yaml to set up the other containers for my project.
version: "3.8"
services:
hadea-database:
image: hadea_oracle_1521:19.3.0
container_name: hadea_oracle_1930
ports:
- "1521:1521"
environment:
- ORACLE_SID=ORCLCDB
- ORACLE_PDB=ORCLPDB
- ORACLE_PWD=Oracle4System
- ORACLE_MEM=2000
volumes:
- ./database/OracleDB/oradata:/opt/oracle/oradata
- ./database/OracleDB/setup:/opt/oracle/scripts/setup
- ./database/OracleDB/startup:/opt/oracle/scripts/startup
networks:
- hadea-network
hadea-maildev:
image: maildev/maildev
container_name: hadea_maildev
command: bin/maildev --web 80 --smtp 25 --hide-extensions STARTTLS
ports:
- "8081:80"
networks:
- hadea-network
hadea-server:
build:
context: ./server
dockerfile: Dockerfile
container_name: hadea_back
environment:
- HTTP_PORT=3000
- HTTP_HOST=0.0.0.0
- DATABASE_HOST=hadea-database
- DATABASE_PORT=1521 # CONTAINER port NOT the HOST port
- DATABASE_SID=ORCLCDB
- MAIL_HOST=hadea-maildev
- MAIL_PORT=25 # CONTAINER port NOT the HOST port
ports:
- "3000:3000"
working_dir: /usr/src/app
volumes:
- ./server:/usr/src/app
networks:
- hadea-network
depends_on:
- hadea-database
- hadea-maildev
hadea-front:
build:
context: ./front
dockerfile: Dockerfile
container_name: hadea_front
ports:
- "4200:4200"
- "3001:3001"
volumes:
- ./front:/usr/src/app
networks:
- hadea-network
depends_on:
- hadea-database
- hadea-maildev
- hadea-server
networks:
hadea-network:
If you want to change the port used WITHIN the container (I think this is the question), you could try building a new image after modifying the conf file, e.g. (for the 18c image). The other images hard code the 1521 port in various files in that repo depending on the oracle version you are using, so those would have to be changed prior to building the image.
I have been using this image: container-registry.oracle.com/database/express:latest. This is version 18c and it has a conf file within the image located at /etc/sysconfig/oracle-xe-18c.conf, I would just build a new Dockerfile and overwrite that file with a new one that has the port you require. Or, you could extract the entire contents of that directory, dump it to a host directory, modify the file as needed, and map a volume to etc/sysconfig (make sure the permissions are correct). This way you could tweak the file from the host. It might be possible to set the variable in that conf file from an environment variable within a docker-compose.yaml file or on the docker command line. This variable is named LISTENER_PORT. Some of the variables in these scripts are defined locally and do not pull their values from environment variable though.
I have a few microservices running under Docker. They are Zuul, Eureka, and a configuration server. These are working but when I start my authorization-service, it says I cannot connect to PostgreSQL.
version: '3'
services:
eureka-discovery:
...
zuul-gateway:
...
config:
...
postgres:
image: postgres
environment:
POSTGRES_USER: postgres
POSTGRES_PASSWORD: root
POSTGRES_DB: spring_microservices
ports:
- 5432:5432
authorization-service:
image: authorization-service:0.0.1
environment:
- eureka.client.serviceUrl.defaultZone=http://eureka-discovery:8761/eureka
- spring.cloud.config.uri=http://zuul-gateway:8765/config-service
- spring.datasource.url=jdbc:postgresql://postgres:5432/spring_microservices
depends_on:
- eureka-discovery
- zuul-gateway
- config
- postgres
ports:
- 1001:1001
What's confusing is that I am specifying postgres instead of localhost to make the connection, but the error continues to say "localhost". You see I override other properties the same way I do with spring.datasource.url and they work but this one.
org.postgresql.util.PSQLException: Connection to localhost:5432 refused. Check that the hostname and port are correct and that the postmaster is accepting TCP/IP connections.
I attempted a few variations I could find around, such as using a different service name, using SPRING_DATASOURCE_URL, using the connection string as postgresql://[user]:[pass]#[service]/[database], I tried moving this line to the config-service instead since the database connection is indeed in the Spring Cloud Config Server microservice, but all to no avail. What's missing? It seems to be correct comparing to all the solutions I found.
I made Spring Config Server see my repository for the configurations. Thinking better, if it's reading from the repository it's evident that changing it on the fly locally won't work, so I updated the connection string of the configuration file for authorization-service then pushed, and it worked. I wonder if this is a good practice or better to make Config Server see local files and change it on the fly?
Hi i hope you are okay !
above your authorization-service add a link like this:
links:
- "postgres"
Greetings
So, I searched around for an answer on this matter but either people don't address the issue or they say there's no problem doing this on their computer (mac or linux). It seems like this might be a windows problem.
I have a spring api running on a docker container (linux container). I use docker desktop on windows and I'm trying to make a request (in insomnia/postman/wtv) to that api.
If I run the api locally making the following request works perfectly:
http://localhost:8080/api/task/
This will list multiples task elements.
I've containerized this application like so:
Dockerfile
FROM openjdk:11.0.7
COPY ./target/spring-api-0.0.1-SNAPSHOT.jar /usr/app/
WORKDIR /usr/app
RUN sh -c 'touch spring-api-0.0.1-SNAPSHOT.jar'
ENTRYPOINT ["java", "-jar", "spring-api-0.0.1-SNAPSHOT.jar"]
docker-compose.yml
version: '3.8'
services:
api:
build: .
depends_on:
- mysql
environment:
- SPRING_DATASOURCE_URL=jdbc:mysql://mysql:3306/test?createDatabaseIfNotExist=true
ports:
- "8080:80"
mysql:
image: mysql
ports:
- "3306:3306"
environment:
- MYSQL_ROOT_PASSWORD=root
- MYSQL_USER=root
- MYSQL_PASSWORD=root
- MYSQL_DATABASE=test
If I do docker-compose up this works without issue:
The problem is, if I try to call the same endpoint as before from localhost I don't get any response.
Insomnia returns an error saying: Error: Server returned nothing (no headers, no data)
I've also tried connecting to the container's ip (got it from docker inspect) but no luck.
Ports are exposed in docker-compose.yml. What am I missing?
Thanks in advance.
Port mapping is incorrect.Spring boot application started at 8080 (from the image I see) inside container and it should be mapped to 8080 inside the container.
It should be like below:
ports:
- "8080:8080"
I am trying to run Traefik on a Raspberry Pi Docker Swarm (specifally following this guide https://github.com/openfaas/faas/blob/master/guide/traefik_integration.md from the OpenFaaS project) but have run into some trouble when actually trying to connect via https.
Specifically there are two issues:
1) When I connect to http://192.168.1.20/ui I am given the username / password prompt. However the details (unhashed password) generated by htpasswd and used in the below docker-compose.yml are not accepted.
2) Visting the https version (http://192.168.1.20/ui) does not connect at all. This is the same if I try to connect using the domain I have set in --acme.domains
When I explore /etc/ I can see that no /etc/traefik/ directory exists but should presumably be created so perhaps this is the root of my problem?
The relevant part of my docker-compose.yml looks like
traefik:
image: traefik:v1.3
command: -c --docker=true
--docker.swarmmode=true
--docker.domain=traefik
--docker.watch=true
--web=true
--debug=true
--defaultEntryPoints=https,http
--acme=true
--acme.domains='<my domain>'
--acme.email=myemail#gmail.com
--acme.ondemand=true
--acme.onhostrule=true
--acme.storage=/etc/traefik/acme/acme.json
--entryPoints=Name:https Address::443 TLS
--entryPoints=Name:http Address::80 Redirect.EntryPoint:https
ports:
- 80:80
- 8080:8080
- 443:443
volumes:
- "/var/run/docker.sock:/var/run/docker.sock"
- "acme:/etc/traefik/acme"
networks:
- functions
deploy:
labels:
- traefik.port=8080
- traefik.frontend.rule=PathPrefix:/ui,/system,/function
- traefik.frontend.auth.basic=user:password <-- relevant credentials from htpasswd here
restart_policy:
condition: on-failure
delay: 5s
max_attempts: 20
window: 380s
placement:
constraints: [node.role == manager]
volumes:
acme:
Any help very much appreciated.
Due to https://community.letsencrypt.org/t/2018-01-09-issue-with-tls-sni-01-and-shared-hosting-infrastructure/49996
The TLS challenge (default) for Let's Encrypt doesn't work anymore.
You must use the DNS challenge instead https://docs.traefik.io/configuration/acme/#dnsprovider.
Or waiting for the merge of https://github.com/containous/traefik/pull/2701