[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Publicity] Blog post: OSSTest Standalone Mode Step by Step

On 26/09/13 13:53, Konrad Rzeszutek Wilk wrote:
On Thu, Sep 26, 2013 at 10:50:02AM +0100, George Dunlap wrote:
On 25/09/13 21:58, Konrad Rzeszutek Wilk wrote:
On Mon, Sep 23, 2013 at 02:17:03PM +0100, Wei Liu wrote:
On Mon, Sep 23, 2013 at 12:34:04PM +0100, Lars Kurth wrote:

thanks. I will leave it to others to comment. Would you be willing to put
together a wiki page based on the post and post to xen-devel? I had a
Yes, sure.

conversation with IanC at LinuxCon about the lack of people using OSS Test.
I believe the two main reasons for the lack of pick-up so far are:
a) It is unclear how to build OSS Test into the dev routine (this post +
wiki should address this)
I think it can address part of the problem, but not all of it. There're
still lots of details I'm not able to go into in a single post.

The sole purpose of this post is to help people bootstrap using OSSTest.
Once they're en route they're pretty much on their own and everything
boils down to "read the source code".
It does detract from the Xen RT which long-term ought to be used (IMHO).

As it has a ton of features that can help in making Xen and Linux
work fantasticly well  - and catch not only bootup/bug regressions
but also performance regressions.
With due respect to Alex and the XenRT team for the work they've
done, XenRT is not actually a public project yet.  Tossing a tarball
over the wall is not making an open project.  And I haven't looked
at it, but I suspect it's not in a shape ready to just be picked up
and used by the average developer just yet.

The thinking here was:
* It may be 6 months to a yeas before XenRT is ready to replace osstest
* There is a small chance that the XenRT thing may not actually come about
I hope not.
* Work done and skills gained writing tests for osstest wouldn't be
wasted -- having implemented tests in osstest should make it easier
for people to implement the same test (or different tests) in XenRT.

So we were planning on continuing to develop osstest until XenRT was
ready to take over, as the risk of doing so is low and the benefits
are solid.

It occurs to me that for the osstest->XenRT transition, a first step
would be to get XenRT running on the xenproject.org testbed. That
will help shake out all the random assumptions about the environment
that may have ended up baked into the code.  We should probably try
to touch base with the XenRT team about that, after IanJ comes back
from holiday.
I completly get the short term - we need osstest, long term lets look
at Xen RT. Perhaps the blog can mention Xen RT as the more longterm
transition and enumerate some of the things it does?

Well the purpose of the blog is a how-to of how to get osstest up and running; having a big section on XenRT would be out of place. Also, I doubt Wei has any personal experience with XenRT, so he's not exactly in a position to extol its benefits. That would be better done by someone on the XenRT team, and in a different blog.

But we should probably at least mention it, and how osstest relates to it. How about something like this:

"Some readers may recall the recent announcement of open-sourcing XenRT, and wondering about the relationship between osstest and XenRT. Long-term, we expect that XenRT will mature as an open development project and eventually displace osstest. But that is not likely to happen for another six months to a year. Developing osstest has benefits for the current development, and we hope that getting people involved in writing test cases for osstest will be of service in helping people write test cases for XenRT when it becomes available. So we have decided to continue to develop osstest until XenRT is ready to replace it."


Publicity mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.