Open sivamind opened 3 years ago
@oshmyheliuk , @NadiyaS , please review.
@sivamind Hi,
Do you still face this error?
Could you please paste the content of your docker-compose.yaml
with code markdown format:
```
code
```
@oshmyheliuk We've received this 504 error for a couple of times as well. One thing that I've seen is the fact the the reverse proxy settings in tls have no timeout set, thus nginx uses the default value. Maybe it's worth tweaking this.
Also, for some reason the timeouts on different containers are so "different" in values
I think we should pass the same value to all parties involved, or decrease it until we reach php, that way if php time out is reached, we can get a response up till tls..
This is what happened to me. Verify all your caches are turned on.
I ran into this also. After deploying (2.3.5-p2) the caches were setup like this:
Current status:
config: 0
layout: 1
block_html: 1
collections: 0
reflection: 0
db_ddl: 0
compiled_config: 0
config_integration: 0
config_integration_api: 0
full_page: 1
target_rule: 0
config_webservice: 0
translate: 0
vertex: 0
I have the same issue also, 504 quite often
version: '2.1'
services:
mysql_magento:
image: 'mariadb:10.2'
environment:
- MYSQL_ROOT_PASSWORD=magento2
- MYSQL_DATABASE=magento2
- MYSQL_USER=magento2
- MYSQL_PASSWORD=magento2
volumes:
- './.docker/mysql/:/var/lib/mysql'
healthcheck:
test: 'mysqladmin ping -h localhost'
interval: 30s
timeout: 30s
retries: 3
networks:
magento:
aliases:
- mysql_magento
elasticsearch_magento:
image: markoshust/magento-elasticsearch:7.6.2-2
environment:
- "discovery.type=single-node"
networks:
magento:
aliases:
- elasticsearch_magento
fpm:
hostname: fpm.magento2.docker
image: 'magento/magento-cloud-docker-php:7.3-fpm-1.1'
extends: generic
volumes:
- '.:/app:delegated'
- './.docker/mnt:/mnt:rw,delegated'
environment:
- 'PHP_EXTENSIONS=bcmath bz2 calendar exif gd gettext intl mysqli pcntl pdo_mysql soap sockets sysvmsg sysvsem sysvshm opcache zip xsl sodium xdebug'
networks:
magento:
aliases:
- fpm_magento
depends_on:
mysql_magento:
condition: service_started
web:
hostname: web.magento2.docker
image: 'magento/magento-cloud-docker-nginx:1.19-1.2.1'
extends: generic
volumes:
- '.:/app:delegated'
environment:
- WITH_XDEBUG=1
- NGINX_WORKER_PROCESSES=4
- NGINX_WORKER_CONNECTIONS=1024
networks:
magento:
aliases:
- nginx_magento
depends_on:
fpm:
condition: service_started
varnish:
hostname: varnish.magento2.docker
image: 'magento/magento-cloud-docker-varnish:6.2-1.2.1'
networks:
magento:
aliases:
- varnish_magento
depends_on:
web:
condition: service_started
tls:
hostname: tls.magento2.docker
image: 'magento/magento-cloud-docker-nginx:1.19-1.2.1'
extends: generic
volumes:
- './.docker/log/:/var/log/nginx/:delegated'
networks:
magento:
aliases:
- tls_magento
environment:
- NGINX_WORKER_PROCESSES=4
- NGINX_WORKER_CONNECTIONS=1024
- UPSTREAM_HOST=varnish
- UPSTREAM_PORT=80
depends_on:
varnish:
condition: service_started
generic:
hostname: generic.magento2.docker
image: 'magento/magento-cloud-docker-php:7.3-cli-1.1'
volumes:
- './.composer:/app/.composer:rw,delegated'
env_file: ./.docker/config.env
environment:
- MAGENTO_RUN_MODE=developer
- 'PHP_EXTENSIONS=bcmath bz2 calendar exif gd gettext intl mysqli pcntl pdo_mysql soap sockets sysvmsg sysvsem sysvshm opcache zip redis xsl sodium'
redis_magento:
image: redis:5.0-alpine
networks:
magento:
aliases:
- redis_magento
mailhog:
hostname: mailhog.magento2.docker
image: 'mailhog/mailhog:latest'
networks:
magento:
aliases:
- mailhog_magento
phpmyadmin:
image: phpmyadmin
environment:
- PMA_HOST=mysql_magento
- PMA_PORT=3306
volumes:
- './.docker/phpmyadmin/sessions:/sessions'
networks:
magento:
aliases:
- phpmyadmin_magento
build:
image: 'magento/magento-cloud-docker-php:7.3-cli-1.1'
extends: generic
volumes:
- '.:/app:rw,delegated'
- './vendor:/app/vendor:rw,delegated'
- './generated:/app/generated:rw,delegated'
- '~/.composer/cache:/root/.composer/cache:rw,delegated'
- './.composer/auth.json:/root/.composer/auth.json:rw,delegated'
networks:
magento:
aliases:
- magecli_magento
depends_on:
mysql_magento:
condition: service_started
redis_magento:
condition: service_started
elasticsearch_magento:
condition: service_started
fpm_xdebug:
hostname: fpm_xdebug.magento2.docker
image: 'magento/magento-cloud-docker-php:7.3-fpm-1.1'
extends: generic
volumes:
- '.:/app:ro,delegated'
- './vendor:/app/vendor:ro,delegated'
- './generated:/app/generated:rw,delegated'
- './var:/app/var:rw,delegated'
- './app/etc:/app/app/etc:rw,delegated'
- './pub/media:/app/pub/media:rw,delegated'
- './pub/static:/app/pub/static:rw,delegated'
- './.docker/mnt:/mnt:rw,delegated'
environment:
- 'PHP_EXTENSIONS=bcmath bz2 calendar exif gd gettext intl mysqli pcntl pdo_mysql soap sockets sysvmsg sysvsem sysvshm opcache zip xsl sodium xdebug'
networks:
magento:
aliases:
- fpm_xdebug.magento2.docker
depends_on:
mysql_magento:
condition: service_started
networks:
magento:
driver: bridge
volumes:
sockdata:
Thank you for reporting his. We have created an internal ticket MCLOUD-7992 to investigate
AFAIK this happens cuz docker + developer mode + local environment + 0 caches. So, there are only 2 workarounds for this
With point 2 you will still get some 504 (during page builder related requests most of the time) occasionally but nothing that a fast refresh cant fix :)
In my case the issue was related to implicitly default 60 second value of proxy_read_timeout in /etc/nginx/conf.d/default.conf incide TLS container.
Solved this by adding proxy_read_timeout 600;
in server listen 80 section under location / where the rest of proxy_pass instructions are set for varnish.
docker exec -ti $(docker ps --format "{{.Names}}" | grep tls) bash
apt install vim
vim /etc/nginx/conf.d/default.conf
below is my yaml file and host system is windows 10 version: '2.1' services: db: hostname: db.magento2.docker image: 'mariadb:10.2' environment: