I'd like to be able to setup a project to never cache the steps that were in place when a release was created. Or at least have a way to update that cache the same way you can update to the latest variable set.

When we update the deployment steps now we have two options:
1) Rebuild from source (which we should never have to do)
2) Delete and re-add the release (which loses any deployment history)

It would be good if Octopus could allow users to add a custom message that any user that logs in can see when the server is in Maintenance Mode. It would even be cooler if an alert message could be scheduled, so admins could add a "Server will be down from X to X" a couple of hours prior to the outage.

Support vor databases other than SQL Server.

Allow cloning of library variable sets so it becomes easier to create a new variable set for a new environment.

I have created a powershell step template to ensure that a package exists at MyGet to work around their eventual consistency ahead of attempting to fetch the package for deployment. In order to utilize this step template, I must also remember to include the script module that it depends upon. It would be great to be able to attach script modules to "run a script" step templates so that you don't have to remember to add the module in addition to the step itself.

Any user that knows the Octopus portal URL can log into Octopus, which will create them a user on the Octopus DB and add them to the "Everyone" group, create entries on the Audit log, etc.

Even though the user won't have any privileges out of the box, implementing a white list system to only allow people in specific AD groups to login would be something good.

Octopus tool to generate Report based on
1. What's Deployed ?
2. Who Deployed ?
3. Deployed To Where ?
4. Who Approved ? (Manual Intervention Step)
5. When did the deployment Take place (Week Day & Weekend deployments)

When I navigate from the dashboard to the Project Overview page, the page should only show the environments that I selected on the Configure Dashboard screen. Also, it would be helpful to filter the list of environments by channel or lifecycle.

When a tentacle is installed on a machine, the recovery on the windows service should be set to restart the service on failure.

I had several tentacle services fail, which caused deployments to fail on those nodes

It would be useful to be able to have better organizational power with Project Groups. Is as much, we should be allowed to employ a hierarchical structure with projects spanning many levels deep (instead of one level deep currently with the Project Group > Project model currently used). Look at how TeamCity organizes builds within projects as an example.

This would allow for potential further enhancement say with cascading project group security, environment, variable set, script module, and lifecycle etc inheritance. Again, look to the TeamCity model for more on what I mean...

When working with a number of library variables, sometimes we don't name them as well as we should. Inevitably, we look at a variable and say "where is this used again? Client side transforms? A step template?" -- If Octopus was able to report which variables were used where, it could help us better remove unused variables, and help us better organize the existing ones.

Some of our ops need different dashboard views depending on what they are interested in at the time ... would be could to be able to save 'named' dashboard views that you can easily change in a drop down menu or something similar.

We happened to be hosting https://demo.octopusdeploy.com and https://demo.octopus.com using two difference certificates, but the Octopus Manager really only supports a single SSL binding - the last one in the script wins.

Here's an example of the script that worked using hostnameport instead of ipport:

"netsh.exe" http delete sslcert ipport=0.0.0.0:443
"netsh.exe" http add sslcert hostnameport=demo.octopus.com:443 appid={E2096A4C-2391-4BE1-9F17-E353F930E7F1} certhash=FAKE certstorename=My
"netsh.exe" http add sslcert hostnameport=demo.octopusdeploy.com:443 appid={E2096A4C-2391-4BE1-9F17-E353F930E7F1} certhash=FAKE certstorename=My
"C:\Program Files\Octopus Deploy\Octopus\Octopus.Server.exe" configure --instance "OctopusServer" --webForceSSL "True"
"C:\Program Files\Octopus Deploy\Octopus\Octopus.Server.exe" configure --instance "OctopusServer" --webListenPrefixes "http://demo.octopusdeploy.com/,http://demo.octopus.com/,https://demo.octopus.com/,https://demo.octopusdeploy.com/"
"C:\Program Files\Octopus Deploy\Octopus\Octopus.Server.exe" service --instance "OctopusServer" --stop --start

Nuget push is not sufficient for moving my packages from my build environment up to my octopus server, and would prefer to use FTP.

Nuget.Server and Klondike both automatically scan the file system and add new packages to the index as they appear. Octopus’ failure to do so is preventing us from using the built-in feed.

We would like to see more grouping options for Projects, Particularly a nested hierarchical style grouping capability.

Allow email (or other form of) communication when an environment target becomes unhealthy. I ran into a situation where deployments failed because the target was unavailable and I had no idea until the deploy was to occur. This could be a Process Step, but would be nice to proactively know that a server is unavailable ASAP, ya' know?

At the moment there is no way to define order for projects or project groups.

This would be handy for all sorts of reasons but most significantly on the dashboard

When updating Variables, it would be great to see what Octopus will change the web.config to when deployed to an environment without actually deploying it.

This is only a small CSS change but if you could increase the width of the "cell" that contains the project name on the dashboard that would be great. We have lots of screen space but project names (and group names) are wrapping after just 20 odd characters.