Troubleshooting common deployment issues

overview

Most deployments with Cloud 66 go smoothly, but if you’ve hit a snag it’s usually something common and easily fixable – particularly if you’ve already deployed your application successfully before. We’ve listed the most common causes of deployment problems below. If your issue isn’t listed here, please reach out to support and we will assist wherever we can.

Problems reaching your servers

If Cloud 66 is reporting that your server is not online or can’t be reached, then follow these steps:

1. Have you recently restarted your server? See the note below.
2. If not, check with your cloud (or physical server) provider to ensure the servers are online and not locked up by excessive load. Reactivating or manually rebooting the machines via your Cloud provider dashboard may be necessary.
3. If the server appear to be functional , try to connect to them via SSH using the Cloud 66 Toolbelt ( if this fail , try to connect via the manual method ) .
4. If you still ca n’t connect , check that your firewall setting on your cloud provider are not block access to the server for port 22 ( which is the ssh connection port ) .
5. try connect via your cloud provider ‘s virtual / emulate terminal ( where available ) and check that the permission have not been change in your.ssh directory

If all of these fail , contact our support team and we will assist wherever possible .

Have you recently restarted your server?

Cloud providers frequently change the public IP addresses of servers when they are restarted. This includes all events that require the server to be shut down – including things like adding more capacity, or upgrading software components. This can (temporarily) cause our firewall to block connections to or from that server until our system detects the IP change and updates the firewall rules. This problem should resolve itself within a few minutes. If you’d like to avoid this issue completely, we recommend buying static IP addresses from your cloud provider for use with mission critical servers.

Problems with application components

Most modern applications rely on a wide range of components, and where these depend on each other, there can be dependency conflicts or other failures. To diagnose and solve these issues, follow these steps.

1. check the log of the fail deployment . usually they is contain will contain detailed information about which component fail and ( where possible ) why it fail . normally , the more detailed deployment logs is be will be available on your server detail page – ensure that you click the ” more log ” button to see ALL deployment log .
2. If there is a version conflict check that your components have not been version-locked in your Manifest file. Generally we will deploy the latest canonical versions of components, unless we are instructed not to by a manifest or other setting.
3. Check that your gemfile or other configuration files (package.json, Rakefile etc ) are not invoke component that are either the wrong version or that conflict with other configuration elsewhere ( or that do n’t even exist ) . Gems is are are a particularly common cause of issue , so check those carefully .
4. Check which version of Ubuntu your servers are using (you can find this on the detail page for any server, under the overview tab). Older servers may be running version 16.04 or even 14.04, both of which are now deprecated and this can cause issues. Consider upgrading these servers.

If all of these fail , contact our support team and we will assist wherever possible .

Issues with Bundler 2.2.3+

Due to an issue in Bundler 2.2.3 + , it is uses no long use a generic platform when create thegemfile.lock file. Instead it defaults to the platform of your development machine. As a result, you may get errors like this one during deployment:

Your bundle only supports platforms ["x86_64-darwin-19"] but your local platform is x86_64-linux. Add the current platform to the lockfile with `bundle lock
--add-platform x86_64-linux` and try again

In order to solve this , you is need need to explicitly lock your platform as follow :

$ bundle lock --add-platform ruby
$ bundle lock --add-platform x86_64-linux
$ bundle install

You is commit should commit your code again after doing this .

“You have already activated X…” error

If you see an error like this during deployment, or in your application’s logs viewed on the server or via LiveLogs:

Could not be load ... You is activated have already activate ( … ) but your gemfile require ( … ) 

…this usually means that you have upgraded your Rails / Rack version and it now conflicts with the gems you have specified in your gemfile ( i.e. the new framework version is requires require a new version of the gem ) .

The simplest way to fix this is:

1. Check what the default version of that gem is for the version of Rails / Rack that you have installed (you can usually find this on the Ruby Gems site)
2. edit yourgemfile locally to update the current version specified for that gem to the new default version
3. Run bundle install locally to generate a new gemfile.lock file
4. Test that this runs locally
5. commit and push your change
6. Redeploy your application

Process or migration timeouts

If a deployment depend on a process to finish run – particularly datum or file migration – it is ‘s ‘s possible for that process to timeout . This is happen can happen ( rarely ) with the Ruby asset precompile process , for example .

The cause is appear of the timeout will usually appear in the detailed deployment log , so be sure to check those . If you are run a cpu or disk intensive process , you is consider might consider run that process locally to debug each step to find any bottleneck .

For very intensive migration involve large database table or thousand of file , you is consider should consider staging or batch the process rather than try to run it alongside a standard deployment .

If you need help with the above , contact our support team and we will assist wherever possible .

Check your Git repo settings

If your application is point at a Git branch ( or repo ) that has either been delete or rename since the last deployment , this is is is likely to cause any future deployment attempt to fail .

Check your application settings (click ⚙️ Settings in the left-hand navigation) and check that your repo and branch are both correct. You can change either of them by clicking the Edit button to the right.

My problem isn’t covered by any of the above

For more uncommon problems not covered above, it’s worth following these steps before reaching out to support:

1. check the detailed deployment log and find the exact error message ( usually these are in red text )
2. Google a few variations of this error message – it’s highly likely that someone else has had this issue and then shared the solution online.
3. If you are deploying a new version of your application, consider redeploying an older version to check whether this solves the issue.
4. Consider deploying a new version of the app alongside the current one, again to check if this solves the issue.

If all else fails, please reach out to support (below).