View logs for troubleshooting

This topic discusses how you can use logs to troubleshoot problems related to Magento Commerce build hooks and deploy hooks.

You may need to SSH into the environments to locate and view logs.

  • Integration: Copy the link for the SSH command per environment through the Project Web Interface
  • Staging: ssh -A <project ID>_stg@<project ID>
  • Production: ssh -A <project ID>@<project ID>

Build logs

After pushing to your environment, you can see the results of the both hooks. Logs from the build hook are redirected to the output stream of git push, so you can observe them in the terminal or capture them (along with error messages) with git push > build.log 2>&1.

Deploy logs

You can review these logs via SSH into the environment. Change to the directories listed below to review the logs.

Logs from the deploy hook are located on the server in the following locations:

  • Integration: /tmp/log/deploy.log
  • Staging: /var/log/platform/<prodject ID>/post_deploy.log
  • Production: /var/log/platform/{1|2|3}.<prodject ID>/post_deploy.log

The value of <project ID> depends on the project ID and whether the environment is Staging or Production. For example, with a project ID of yw1unoukjcawe, the Staging environment user is yw1unoukjcawe_stg and the Production environment user is yw1unoukjcawe.

For example, on the Staging environment for project yw1unoukjcawe, the deploy log is located at /var/log/platform/yw1unoukjcawe_stg/post_deploy.log.

For Production, you have a three node structure. Logs are available with specific information for that node. For example, on the Production environment for project yw1unoukjcawe, the deploy log is located at node 1 /var/log/platform/1.yw1unoukjcawe/post_deploy.log, node 2 /var/log/platform/2.yw1unoukjcawe/post_deploy.log, and node 3 /var/log/platform/3.yw1unoukjcawe/post_deploy.log.

Logs for all deployments that have occurred on this environment are appended to this file. Check the timestamps on log entries to verify and locate the logs you want for a specific deployment.

The actual log output is highly verbose to allow troubleshooting. The following is a condensed example:

[2016-10-11 22:15:38] Starting pre-deploy.
[2016-10-11 22:15:39] Pre-deploy complete.
[2016-10-11 22:15:42] Start deploy.
[2016-10-11 22:15:42] Preparing environment specific data.
[2016-10-11 22:15:42] Initializing routes.

... more ...

[2016-10-11 22:15:46] Deployment complete.

The deploy log contains start and stop messages for each of the two hooks: Starting pre-deploy, Pre-deploy complete., Start deploy., and Deployment complete..

Application logs

To review other application logs in Staging or Production, you can access and review those logs in /var/log/platform/ProjectID.

For Staging, the project ID has _stg at the end. For example, if you receive 500 errors in Staging and want to review the nginx logs, you can SSH to the Staging environment and locate the logs in /var/log/platform/ProjectID_stg.

Remember, when viewing the logs for Production, you have three nodes to check.