Walking in the shadows of giants

As I sit here in 22A on an American Airlines flight from San Francisco to O'Hare at the start of my 16 hour journey home to Ireland, I'm reflecting on some of the key Solaris 11 related events at Oracle OpenWorld this week.

For the first time in a couple of years, I got to spend the weekend in Northern California, having been here  last week for Solaris 11 planning meetings.  I went up to the Sierras to hug some Sequoias.  I'm not normally the tree-hugging type, but I make as exception for these giants.  I saw Mono Lake.  Cool.  Devil's Postpile.  Way Cool.  And the Sequoia National Park - it's truly amazing walking in the shadows of these giants.

As usual, Oracle OpenWorld and Jave One this week provided the opportunity to hear about bleeding edge technologies directly from their architects and to chat with them about the what and the why.

Markus Flierl (VP, Solaris Engineering) hosted a session on Monday with some of his key architects who have been developing Solaris 11 over the last 7+ years, including Liane Praza (IPS), Bart Smaalders (IPS), Darren Moffett (Security), Dan Price (Zones), and Mark Maybee (I/O).  It was great to hear these experts express their passion, ingenuity, and innovation.  They have a justifable parental sense of pride in Solaris 11.  Technologies which were bolt-ons in Solaris 10, or indeed far too disruptive to even be considered for release in a Solaris 10 Update, are tightly integrated and honed in Solaris 11.  Low latency (i.e. performance), scalability, security, availability, robustness, and diagnosability are all factors that customers have come to expect of Solaris.  Solaris 11 takes it to a whole new level.  Warp drive.

My colleague, Pete Dennis, and I have been working closely with Bart, Liane, David Comay, and others to ensure that IPS fully meets the needs of our customers' maintenance lifecycle.  They've listening to us and subtly tweaked and adapted their implementations where necessary to fully meet customers' maintenance lifecycle needs.  Working with geniuses is great.  Working with geniuses who are prepared to listen and adapt is truly wonderful.

But what really blew me away this week was a presentation by Nicolas Droux last night on Network Virtualization in Solaris 11.  Some of you may know about earlier incarnations of this, codenamed Project "Crossbow".  But the fleshing out of the capabilities in Solaris 11 is truly amazing.  The ability to have virtualized NICs (VNICs), virtualized LANs (VLANs), Zones which act as virtualized switches, Zones which act as virtualized firewalls, fully segregated data "Lanes", "Flows", etc., etc., and all with diagnosability built in with new utilities such as 'dlstat' (Data link stats), 'flowstat', etc.  I hadn't met Nicolas before but wow!  Not only is Nicolas a key architect, he has an amazing ability to explain it with crystal clarity in a really easy to understand manner.  As I said to the Product Manager, Joost Pronk, we've got to video Nicolas giving this talk once Solaris 11 ships so that the world can see it.  

At the end of Nicolas's presentation, Thierry Manfe showed how he is leveraging Network Virtualization in Oracle Solaris's cloud infrastructure provided to enable ISVs to test their apps with complete data integrity and segregation.  You can sign up for this, it's available now.  "Solaris 11. #1 for Clouds" isn't just some Marketing hype. It's true.

I'm walking in the shadow of giants.  And it's a wonderful feeling.

Roll on Solaris 11.  It won't be long now and I really can't wait.  It's amazing.  Big time!

Thank you to the 90+ of you who attended Pete Dennis, Isaac Rozenfeld, and my presentation on Solaris 11 Customer Maintenance Lifecycles, policies, and best practices.  If you missed it, there'll be another chance to catch an updated version with more technical content at DOAG (the German Oracle Users Group) conference in Nuremberg, Germany in November (see previous posting for details).

Finally, I'd like to pay my respects to a true giant of our industry, Steve Jobs.  Gone way too soon.  RIP Steve.  You'll be missed.  Big time!

Best Wishes,

Gerry.

Disclaimer: Any forward looking statements in this posting are subject to the vagueries of my Crystal ball, possible hallucinations, and lack of coffee.  You get the drift. 

Comments:

"Working with geniuses is great. Working with geniuses who are prepared to listen and adapt is truly wonderful."

Prepared to listen? I have practically begged, pleaded, pointed out numerous times that IPS not allowing scripting ("a no scripting zone") and not being able to trigger events in SMF (during removal, for instance), makes IPS unusable for deploying configuration packages.

What did all that effort amount to? Absolutely nothing.

Posted by UX-admin on October 12, 2011 at 02:51 PM IST #

Hi "UX-admin"

I'll pass your comments on to the relevant folk and I'm happy to push them if you can provide details of exactly what your requirement are directly to me off-line (Gerry.Haskins@oracle.com).

However, as I pointed out in my Oracle OpenWorld presentation on the Solaris 11 Customer Maintenance Lifecycle, getting rid of free format patch scripting is a key benefit of IPS, as patch scripts were a recurring source of error as they are a bugger to test in all possible install scenarios. Personally, from bitter experience, I see patch scripts as the work of the devil.

I know you are talking about scripts at a different level, which Sys Admins can tailor and hone to their specific environments, but having all actions under SMF does allow some pretty neat leveraging for higher level enterprise administration tools in future releases.

BTW: I am definitely not an IPS expert yet. I'm still working my way through "man pkg" and lots of other documentation to understand how it all hangs together. What I do know it that the experience in Solaris 11 is a lot more coherent and integrated with other technologies that it was in OpenSolaris or Solaris 11 Express. There's still much more enhancements to be done, but it's looking pretty good IMHO.

Best Wishes,

Gerry.

Posted by guest on October 18, 2011 at 02:18 PM IST #

Did you manage to get anywhere with this issue Gerry? I totally agree with UX-admin regarding IPS not allowing scripting so it would be interesting to find out if you managed to make any progress?

Posted by Johnny on October 11, 2012 at 09:11 AM IST #

Hi Johnny,

Firstly, apologies for my delay in responding - SPARC SuperCluster has been keeping me rather busy! :)

While as a former Sys Admin myself, I fully understand the desire to have full control by being able to tweak installation scripts, the IPS / Solaris folks remain of the opinion that that opens up a Russian Roulette which is not in our customers' interest.

With my patch / SRU QA hat on, I know my team used (abused ?) the patch scripts to workaround various issues, particularly in Solaris 10 Kernel patches.

However, I also know that patch scripts written by development engineers who hadn't considered all the possible installation permutations - e.g. alternate boot env, miniroot, diskless client - were a recurring source of issues.

I fully expected the IPS / Solaris folks to have to row back on their "no scripting" mandate as the realities of the real world impinged upon the neatness of their theoretical model, but I must admit to being rather pleasantly surprised by how robust the IPS implementation in Solaris has proved to be.

By all means continue to let me and, more importantly, the IPS alias / community know about evidence of deficiencies in the current capabilities to manage real world situations.

But for now, from what I can see, it does indeed appear to be fit for purpose.

Best Wishes,

Gerry.

Posted by guest on April 15, 2013 at 11:55 AM IST #

Post a Comment:
  • HTML Syntax: NOT allowed
About

This blog is to inform customers about patching best practice, feature enhancements, and key issues. The views expressed on this blog are my own and do not necessarily reflect the views of Oracle. The Documents contained within this site may include statements about Oracle's product development plans. Many factors can materially affect these plans and the nature and timing of future product releases. Accordingly, this Information is provided to you solely for information only, is not a commitment to deliver any material code, or functionality, and SHOULD NOT BE RELIED UPON IN MAKING PURCHASING DECISIONS. The development, release, and timing of any features or functionality described remains at the sole discretion of Oracle. THIS INFORMATION MAY NOT BE INCORPORATED INTO ANY CONTRACTUAL AGREEMENT WITH ORACLE OR ITS SUBSIDIARIES OR AFFILIATES. ORACLE SPECIFICALLY DISCLAIMS ANY LIABILITY WITH RESPECT TO THIS INFORMATION. ~~~~~~~~~~~~ Gerry Haskins, Director, Software Lifecycle Engineer

Search

Categories
Archives
« April 2014
MonTueWedThuFriSatSun
 
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