FCS Quality All the Time

The ON organization talks a lot about "FCS Quality All the Time".  That focus is the reason for much of its software development process and is mentioned often in discussions, advice to new developers and ON documentation such as "The Quality Death Spiral".

This focus is also a reason nightly ON builds are run on at least one internal engineering server.  This sometimes causes internal pain when something goes wrong, but the production testing exposes problems that will not be encountered at customer sites.  And it's a terrific learning experience if your putback causes a problem :-).

So what does this have to do with my team?  We run Nevada builds on the machines that support the opensolaris.org website.

Yes, development builds on external-facing production machines.  I get interesting responses when people hear about this or figure it out for the first time.  Some responses are simply dropped jaws and looks of incredulity.  Some are comments such as 'are you crazy?' or 'why in the world would you do that?'.

When we investigated self-hosting before the initial launch, we knew we wanted to run Solaris on the machines that supported the OpenSolaris program.  Doing anything else just didn't make sense.

But then we decided it would be better to run Nevada builds because we wanted to make use of new functionality not in Solaris releases at the time (e.g., ZFS) and new features related to existing technologies like Zones and SMF.

And it's worked.  From August 2006 until around March of this year, we ran Build 45 on the infrastructure machines (note that the current build in progress is Build 100).  We finally hit some system problems this Spring that were fixed in later ON builds, so we started upgrading.  And in general, we want to set up a regular rolling upgrade plan to keep more recent builds on these machines.

But a development build worked for more than a year and a half.  Without system problems.  On external-facing production machines that support a public website.  Perhaps I'm easily amused, but that's amazing to me.

And, I think,  a tribute to the focus on "FCS quality all the time".  So thanks to the ON organization for its diligence about this focus.  It clearly makes a difference.


P.S. Yes, we plan to move to OpenSolaris builds at some point in the future.  But at the moment, our list is long and the product is new, so planning for that transition is not in the works yet.

Comments:

[Trackback] Opensolaris has the reputation of being the unstable tree of Solaris. Nowhere near of production quality. Like the development builds of Linux. But the reality is: There is nothing like a experimental tree in Solaris. When you want to do development yo...

Posted by c0t0d0s0.org on September 26, 2008 at 06:06 AM MDT #

Why would you even want to move to OpenSolaris, what's wrong with Solaris Express CE?

The whole "Ian & Co." marketing tactic is really frustating. All of a sudden, "OpenSolaris" is the new religion, for no other reason than marketing.

Please stick with Solaris Express CE, and don't believe the hype!

Posted by UX-admin on September 26, 2008 at 09:22 PM MDT #

Not that I'm doubting your math, but August 2006 to March 2008 is more likely 20 months than 8.

Posted by Ceri Davies on September 27, 2008 at 12:08 AM MDT #

Hi Ceri - You should definitely doubt my math - thanks for the good catch :-) I have no idea where I got 8 months but I was probably doing multiple things at the same time. Entry is updated to note the more accurate "more than a year and a half". Thanks!

Posted by Bonnie Corwin on October 28, 2008 at 05:07 AM MDT #

Post a Comment:
  • HTML Syntax: NOT allowed
About

Bonnie manages the OpenSolaris Developer Collaboration team that owns and supports the opensolaris.org website. She lives in Colorado, hence the picture of Pikes Peak taken by her husband, Eric Corwin.

Search

Categories
Archives
« April 2014
SunMonTueWedThuFriSat
  
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
   
       
Today