Skip to content

Latest commit

 

History

History
402 lines (313 loc) · 10 KB

09-troubleshooting.md

File metadata and controls

402 lines (313 loc) · 10 KB
Raw

Troubleshooting

This document contains solutions to the most common issues related to the Docker SDK.

Troubleshooting installation

This section describes common issues related to installation.

Docker daemon is not running

when Running the docker/sdk up console command returns a similar error:

Error response from daemon: Bad response from Docker engine

then

  1. Make sure Docker daemon is running.
  2. Run docker/sdk up again.

docker-sync cannot start

when Running docker-sync clean returns an error similar to the following:

docker: Error response from daemon: Conflict. The container name "/data-sync" is already in use by container "47dd708a7a7f9550390432289bd85fe0e4491b080748fcbba7ddb3331de2c7e7". You have to remove (or rename) that container to be able to reuse that name.

then

  1. Run docker-sync clean.
  2. Run docker/sdk up again.

when You get an error similar to the following:

Unable to find image "eugenmayer/unison:hostsync_@.2' Locally
docker: Error response from daemon: manifest for eugenmayer/unison:hostsync_@.2 not found: manifest unknown: manifest unknown.

then Update docker-sync:

gem install docker-sync

Setup of new indexes throws an exception

when Running the command setup-search-create-sources [vendor/bin/console search:setup:sources] returns the exception:

Elastica\Exception\Connection\HttpException - Exception: Couldn't resolve host
in /data/vendor/ruflin/elastica/lib/Elastica/Transport/Http.php (190)

then Increase RAM for Docker usage.

Vendor folder synchronization error

when You get an error similar to vendor/bin/console: not found.

then Re-build basic images, assets, and codebase:

docker/sdk up --build --assets

An error during front end setup

when The frontend:project:install-dependencies command returns an error similar to the following:

-->  DEVELOPMENT MODE
Store: US | Environment: docker
Install Project dependencies
[info] npm
[info]  WARN prepare
[info]  removing existing node_modules/ before installation
[info]
> fsevents@1.2.11 install /data/node_modules/fsevents
> node-gyp rebuild
[info] gyp
[info]  ERR! find Python
gyp ERR! find Python
[info]  Python is not set from command line or npm configuration
gyp ERR!
[info] find Python Python is not set from environment variable PYTHON
gyp ERR!
[info]  find Python checking if "python" can be used
gyp ERR!

then

  1. In deploy.*.yaml, change the base PHP image:
image:
    tag: spryker/php:7.3-alpine3.10
  1. Fetch the changes and start the instance:
docker/sdk boot && docker/sdk up

Demo data was imported incorrectly

when Demo data was imported incorrectly.

then Re-load demo data:

docker/sdk clean-data && docker/sdk up --data && docker/sdk console q:w:s -v -s

Troubleshooting running applications

This section describes common issues related to running applications.

Port is already occupied on host

when Running the docker/sdk up console command returns an error similar to the following:

ERROR: for nginx_frontend Cannot start service nginx_frontend: driver failed programming external connectivity on endpoint spryker_nginx_frontend_1 (e4fdb360f6c9a3243c0a88fa74f8d377325f65b8cd2340b2dacb51377519c1cf): Error starting userland proxy: Bind for 0.0.0.0:80: unexpected error (Failure EADDRINUSE)

then

  1. Check what process occupies the port:
sudo lsof -nPi:80 | grep LISTEN
  1. Stop the process or make it use a different port.
  2. Run docker/sdk up again.

413 Request Entity Too Large

when You get the 413 Request Entity Too Large error.

then

  1. Increase the maximum request body size for the related application. See Deploy File Reference - 1.0 to learn how to do that.
  2. Fetch the update:
docker/sdk bootstrap
  1. Re-build applications:
docker/sdk up

Nginx welcome page

when You get the Nginx welcome page by opening an application in the browser.

then

  1. Update the nginx:alpine image:
docker pull nginx:alpine
  1. Re-build applications:
docker/sdk up

An application is not reachable via http

when An application like Yves, BackOffice(Zed), GlueStorefront(Glue), GlueBackend or MerchantPortal is not reachable after installation.

then In deploy.*.yml, ensure that SSL encryption is disabled:

docker:
    ssl:
        enabled: false

Mutagen

when You get the error:

unable to reconcile Mutagen sessions: unable to create synchronization session (spryker-dev-codebase): unable to connect to beta: unable to connect to endpoint: unable to dial agent endpoint: unable to create agent command: unable to probe container: container probing failed under POSIX hypothesis (signal: killed) and Windows hypothesis (signal: killed)

then

  1. Restart your OS.
  2. If the error persists: Check Mutagen documentation.

when There is a synchronization issue.

then

  • Restart your OS.
  • Run the commands:
docker/sdk trouble
mutagen sync list
mutagen sync terminate <all sessions in the list>
docker/sdk up

when Revert on specific Mutagen version. E.G you are using Docker Compose V1 and you don't have the possibility to update to the docker compose v2 (mandatory requirement).

then

brew uninstall --ignore-dependencies {{ mutagen || mutagen-beta }}
  • Run the commands:
cd "$(brew --repo mutagen-io/homebrew-mutagen)" && \
git checkout {{ HASH COMMIT FROM mutagen-io/homebrew-mutagen }} && \
HOMEBREW_NO_AUTO_UPDATE=1 brew install mutagen-io/mutagen/{{ mutagen || mutagen-beta }} && \
mutagen daemon stop  && \
mutagen daemon start && \
cd -

when Error:

unable to bring up Mutagen Compose sidecar service: unable to reconcile Mutagen sessions: unable to connect to Mutagen daemon: client/daemon version mismatch (daemon restart recommended)

then

  • Run the commands:
mutagen daemon stop
docker/sdk prune

Troubleshooting debugging

This section describes common issues related to debugging.

when Xdebug does not work.

then

  1. Ensure that Xdebug is enabled in deploy.*.yml:
```yaml
docker:
...
    debug:
      xdebug:
        enabled: true
  1. Ensure that IDE is listening to the port 9000.
  2. Check if the host is accessible from the container:
docker/sdk cli -x bash -c 'nc -zv ${SPRYKER_XDEBUG_HOST_IP} 9000'

when nc command does not give any output.

then Contact us.

when nc command tells that the port is opened.

then

  1. Check what process occupies the port 9000 by running the command on the host:
sudo lsof -nPi:9000 | grep LISTEN
  1. If it's not your IDE, free up the port to be used by the IDE.

when PHP xdebug extension is not active in CLI.

then Exit the CLI session and enter it with the -x argument:

  • docker/sdk cli -x
  • docker/sdk testing -x

when PHP xdebug extension is not active when accessing the website via a browser or curl.

then

Try the following:

  • Set the XDEBUG_SESSION=spryker cookie for the request. You can use a browser extension like Xdebug helper.
  • Run the following command to switch all applications to debug mode: 
    docker/sdk run -x

when It's not possible to establish the database connection from the host machine.

then

  1. Check that the deploy yml file is used and make sure the port is exposed.
  2. Check that port is not occupied by the local process by running sudo lsof -nP -i4TCP:3306 | grep LISTEN (with port declared in the deploy.yml file).
  3. Check if the proper credentials and database name are used. You can find all the required information in the deploy yml file.

when You get an error after running docker/sdk cli {ARGUMENT_1}.

then Wrap the command arguments into single quotes. For example, docker/sdk cli 'composer require spryker/*'

when Node Sass does not yet support your current environment: Linux Unsupported architecture (arm64) with Node.js

then

  1. remove node-sass dependencies in package.json
  2. add sass and sass-loader
...
"sass": "~1.32.13",
"sass-loader": "~10.2.0",
...
  1. update @spryker/oryx-for-zed
...
"@spryker/oryx-for-zed": "~2.11.5",
...
  1. add option to sass-loader (frontend/configs/development.js)
loader: 'sass-loader',
options: {
   implementation: require('sass'),
}
  1. run docker/sdk cli
  2. run npm install to update package-lock.json and install dependencies
  3. (if yarn usage) run yarn install to update package-lock.json and install dependencies
  4. run npm run yves to rebuild yves
  5. run npm run zed to rebuild zed

when Error 403 No valid crumb was included in the request

then Check your project configuration. Jenkins CSRF protection should be enabled.

...
$config[SchedulerJenkinsConstants::JENKINS_CONFIGURATION] = [
    SchedulerConfig::SCHEDULER_JENKINS => [
        SchedulerJenkinsConfig::SCHEDULER_JENKINS_CSRF_ENABLED => true,
    ],
];
...

You can use SPRYKER_JENKINS_CSRF_PROTECTION_ENABLED env variable. This variable depends on deploy file parameter from scheduler

services:
  scheduler:
    csrf-protection-enabled: { true | false }

...
$config[SchedulerJenkinsConstants::JENKINS_CONFIGURATION] = [
    SchedulerConfig::SCHEDULER_JENKINS => [
        SchedulerJenkinsConfig::SCHEDULER_JENKINS_CSRF_ENABLED => (bool)getenv('SPRYKER_JENKINS_CSRF_PROTECTION_ENABLED'),
    ],
];
...

when Assets install error

line 1: run-s: not found

then

  1. ensure you are using latest docker-sdk and spryker/php image.
  2. run docker/sdk pull
  3. run docker/sdk boot {{your deploy file}} && docker/sdk up --build