Currently from the Rightscale security list you either need to completely lock it down or completely open it up and there is no in between. I need something that is more granular for example:
1) Give only permission for a user to modify inputs and run a script.
2) Give permission to only specific deployments
3) Give permission to only specific servers

Basically we need a matrix that goes through all the Rightscale components such as servers, templates, deployments, security groups, ssh keys, credendial, arrays etc...

You can only change current inputs at the server level. I would like to be able to view and update current inputs (not just next inputs) at the deployment level so that I can run a script across a deployment that uses the new value.

3 ideas:
1. Better output from chef when things go wrong. Right now the output from RightLink is entirely too terse (1 stack frame in backtraces, etc). [DONE w/ RightLink 5.7]

2. Support for chef roles. This would allow a user to specify which chef roles a server template is to fulfill instead of having to include the recipes for each of those roles. [Moved to idea http://bit.ly/zNQb6o]

3. The RightLink/RightScale sandbox is currently limited to 0.8.16. Chef 0.10.0 is currently the newest. [Planned]

Quite often I find myself needing to access the rightscale dashboard from my mobile device(s): iphone, ipad, etc.

The current site works (mostly), but it is heavy and hard to use.

Please add a mobile UI or maybe an app?

Allow user to specify which chef roles a ServerTemplate is to fulfill instead of having to include the recipes for each of these roles.

(This task was broken out from a previous task:
http://feedback.rightscale.com/forums/94975-customer-feedback/suggestions/1788629-additional-chef-support-debugging-roles-chef-0-?ref=title)

I'd like to see a Database Manager for MongoDB so that I can use this as the backend for my application servers.

Current API for server array instances is lacking output for EC2 instance types and pricing (on demand or spot). It's limiting our capabilities to use the API for accounting and planning purposes.
It would be very useful to bring arrays instances describe API output closer to describe API output for normal servers.

Currently operations are impossible to perform on array instances through the API. This would be very useful for cleanly removing array instances, and dramatically empower API interfacing.

I like how RS can bring server logs into the UI, and can be useful. However, this doesn't scale well when you add more and more servers.

It would be great if there were a way to access an aggregate of all logs at the deployment level, not just at a single instance.

This would be useful for frontend application level logging as well as server logging.

Alerts and auto-scaling based on single CPU e.g. CPU-0 cause unnecessary false positives and cost because the other CPUs might be idle.

It would be nice if the default ruby installed is 1.9.2.
Ruby 1.9.2 has been out for a while. Making our scripts to work with 1.8.7 seems like a downgrade. We can barely run rightscripts in 1.9.2 after installing rvm, but we cannot run Chef recipes.

Provide read access to server array inputs (equivalent to <parameter> element on standalone server show output).

This information should be made available both for the next launch (current array settings) and for any running instances when accessing the hrefs included in the api/acct/1/server_arrays/1/instances endpoint. (Note: hrefs currently provided in that endpoint are not valid.)

RightScale should create a non-privileged Unix user on my instances for every authorized RightScale user, so I can preserve the audit trail for my users' actions and limit their privilege on the instance. Without this feature, I must disable server_login for everyone I don't have absolute trust in. This significantly impairs my employees' ability to use RightScale and makes me want to consider other options.

Currently the deployment monitoring graphs will not retain their collapsed or uncollapsed state. We have to constantly click open all the monitoring bars each time we go to the page.

We need an ability to validate inputs prior to launching the server. For e.g. Regular Expression Validation , Date validation , Number validation.

We currently have 150 + deployments and it is so easy to make a mistake and realizing the impact.

When I tag instances in RightScale, it would be nice if the related Amazon EC2 instance shared the same tags.

I know there are issues in adding dashboard monitoring to array servers (i.e., when array scales up or down), but even if the quick monitoring is static, it would be a big help for me to have the ability to add some graphs from the array servers to the quick monitoring side bar.