Launchpad Entry: server-lucid-ec2-images-painpoints
Contributors: Scott Moser, Eric Hammond, Boris Devouge, Jos Bouman, Zaid Al Hamami
This should provide an overview of the issue/functionality/change proposed here. Focus here on what will actually be DONE, summarising that so that other people don't have to read the whole spec. See also CategorySpec for examples.
This section should include a paragraph describing the end-user impact of this change. It is meant to be included in the release notes of the first release in which it is implemented. (Not all of these will actually be included in the release notes, at the release manager's discretion; but writing them is a useful exercise.)
It is mandatory.
This should cover the _why_: why is this change being proposed, what justifies it, where we see this justified.
You can have subsections that better describe specific parts of the issue.
This section should describe a plan of action (the "how") to implement the changes discussed. Could include subsections like:
Should cover changes required to the UI, or specific UI that is required to implement this
Code changes should include an overview of what needs to change, and in some cases even the specific details.
- data migration, if any
- redirects from old URLs to new ones, if any
- how users will be pointed to the new way of doing things, if necessary.
It's important that we are able to test new features, and demonstrate them to users. Use this section to describe a short plan that anybody can follow that demonstrates the feature is working. This can then be used during testing, and to show off after release. Please add an entry to http://testcases.qa.ubuntu.com/Coverage/NewFeatures for tracking test coverage.
This need not be added or completed until the specification is nearing beta.
This should highlight any issues that should be addressed in further specifications, and not problems with the specification itself; since any specification with problems cannot be approved.
BoF agenda and discussion
This session is to discuss painpoints that people have with the ubuntu images. We would like to get feedback on how people think the images can be made better or what things are particularly painful for their use of them.
- Some people are concerned about the size of the kernel (-virtual 26mb vs. EC2 50+mb)
- Does this cause maintenance headaches?
- No good reason to have useless drivers/modules in the kernel
- People want different settings in their kernel
- Need bits compiled in
missing latency top, this is important for figuring out what is going on with a machine
It might be true that hardy kernels do not have CONFIG_LATENCYTOP, but karmic kernels do:
$ grep LATENCY /boot/config-2.6.31-302-ec2 CONFIG_HAVE_LATENCYTOP_SUPPORT=y CONFIG_LATENCYTOP=y
- People want different things in their kernels, it's hard to build them right now
- Jos will try to figure this out
- Has been a request for a 1000Hz kernel, not clear why this is interesting given that we (think we are) run a tickless kernel
- This may be useful to asterisk users
- Ephemeral storage (for /tmp, for example) is different for different kinds of instances (small, medium, etc.)
- Want to be able to specify what to do with ephemeral mounts in config settings (RAID, etc.)
- We *think* RAID improves reliability and performance
- Creating a filesystem at boot time is slow
- Network services from Amazon are expensive (things that feel like a load balancer, router, etc.)
- Would be nice to provide this functionality pre-packaged to save cost and to put control in the administrators hands
- Consider a tasksel option to make setting this kind of thing up very easy
- Do this because it makes it easier for the user, not necessarily because you *need* it
- Would be nice to have load balanced server out of the box
- keepalived is the most featureful and documented solution available now
- Would be value in providing recipes/best practices for common situations on Ubuntu Server.
- This information comes up, but might be hard to discover
- Default configuration should be good, might alleviate this painpoint
- Lots of advantages to having an image that is configured by a configuration management tool, as opposed to prepackaged images
- Disadvantage to configure-at-runtime is speed, can be slow when you need to spawn up an image quickly
- Consider bundling as a way to optimize instance spawn speed
- People that need super-fast boot will rebundle, but we expect this to be a smaller group than those that will be satisfied by configure-at-runtime mechanisms
- configure-at-runtime is deemed to have maintenance benefits
- Can cater for convenience or cater for optimization. Catering for convenience is preferred; people that need to rebundle will do so