Day 1 : White Rim Road : Shafer to Airport

If you’re not already awake by the time you start the decent down Shafer (about 7 miles in); you soon will be. It’s a pretty exhilarating 1500ft descent through alpine switchbacks – if you take it easy on the dusty / gravely corners you will probably make it down in one piece.

The rest of the first day’s ride was fairly flat, fast  and easy – following the natural contour that is the White Rim. We made decent progress despite frequent stops to enjoy the stunning views.

My trusty Garmin 800 had few issues in some of the more sheltered canyons but we clocked up a fairly easy 19 or so miles on the first day.

Day1-map

We arrived at the Airport campground mid-afternoon – which left plenty of time set up camp, play some Petanque / Boules, eat dinner and explore  a little and get to know our fellow riders. But mostly we were trying to get out of the wind, which, by dusk had picked up enough to threaten my trusty antique tent (note RimTours provide much nicer, much newer tents). This is the same tent that my wife and I used to back-pack around Greece and Turkey about 25 years ago.

 

 

The camping gear we bought was :

  • 2 x inflatable roll mats
  • 2 x inflatable pillows
  • 2 x 4-season (ie .very warm) sleeping bags
  • 1 very old – 2-person, 3-season tent
  • 2x headlamps
  • 2x solar chargers

In hindsight – I should have bought a newer tent or rented one from RimTours and a large beech blanket or something to throw on the floor of the tent would have added to the comfort. Aside from the first night – we were warm and dry and slept pretty well.

Read Day 2 – Airport to Murphy

 

Goodbye UP3

  

My most recent Jawbone UP3 replacement finally gave up the ghost. It lasted 17 weeks – better than the previous 3 but failed with the same issue as the other 3. I’m not even going to waste time trying to get a replacement or refund – Jawbone have wasted too much of my time already. Great software; good tech support but shitty hardware.

I’ll probably take a look at the new Apple watch when it comes out later this year but any recommendations for fitness trackers are welcome. Must have heart rate and sleep tracking.

Gone Tubeless

The cold weather this weekend had me looking for a bike related activity that didn’t involve grinding through semi-frozen piedmont clay. I’ve been thinking of turning my 2014 Specialized Epic Carbon Comp tubeless and spending 30 very cold mins. yesterday fixing a pinch puncture on the trail sealed the deal.

The Epic’s tires are Specialized Fask Trak 2Bliss Ready so all that was required was to find the tubeless valves that came with the bike and to buy some sealant and actually make the transition.

I started with the front. Fitted the valve (there’s an obvious right an wrong orientation), put the tire back on, unscrewed the valve and then added sealant via the valve stem with a Stan’s syringe kits. Rolled the sealant around by rotating the wheel through a couple of different axes, fitted he valve back on then tried to inflate the tire using a track pump. Lots of pumping but no seal. Used a 20g CO2 canister  and instantly the tire popped onto the rim and sealed – nice sound. But there was still little air (and a little sealant) coming out of the valve so I had to wiggle and rock it while tightening the nut until it stopped.
The rear went on much quicker – fitted the tubeless valve tight by applying some sealant to the area and rocking it and wiggling it while tightening the nut. Didn’t bother using the syringe to add sealant – just poured some into the bottom of the tire before popping the tire on. Another  C02 canister instantly inflated the tire and popped the tire onto the bead. Just like a pro !

IMG_4216

I inflated both tires to 60 psi using a track pump and witnessed the magical sealant doing it’s job on an old, decent sized hole in the front tire. I rode the bike to the end of the street and back just to circulate the sealant and left it in the sun for an hour or so. Checked the pressure when I came back and everything looked good so deflated both tires to my usual 40psi.

IMG_4215

I did a quick trail ride but didn’t notice any immediate difference – I suspect that’s more to do with the current conditions (ice, mud, clay); looking forward to a longer ride later in the week. Will still be packing a spare tube until I’m confident of the sealant’s ability to fix small punctures. I did notice that my usual 40psi feels rock solid so will likely have to experiment with pressure a bit. I’ll post an update when I’ve ridden some more.

Also what should I carry on the trails for repairs now ? Just sealant and C02 ?

JBoss EAP 7 Beta – Open Source Pioneer

There are a few Open Source technologies and products that have spearheaded the drive of Open Source  into the enterprise and managed to overcome historical objections  – Linux, Apache Web Server, MySQL, Postgres, WordPress, Hadoop, to name some of the better known technologies. Those technologies paved the way for the open source revolution of the last decade; every enterprise vendor and every organization has adopted open source to some degree. Open Source has won; get over it.

I think Red Hat’s JBoss EAP  (upstream WildFly , previously JBoss AS) is one of those pioneering technologies. It’s hasn’t merely broken through into the enterprise but broken into a strategically important segment that was previously dominated by tech. giants such as Oracle and IBM. It was a segment serviced by Mainframes, Transaction Processing Monitors, Object Request Brokers and highly proprietary 4GLs. As a side note – those same companies are still peddling proprietary products and technologies to solve those same problems.

Building on the early success of Java SE, Java EE was introduced back in 1999 as a collaboration between Sun Microsystems, Oracle and IBM. It was designed as an Open Standard platform for creating complex business applications. At the time, the unlikely collaborators all had a common enemy – Microsoft  – who had spent the last 15 years dominating the desktop with their brand of proprietary technology and were starting to make inroads into the data center. How times have changed.

The excitement around Java, the Open collaboration (through the JCP) and the large ecosystem of vendors supporting the standard helped Java EE dislodge previous generations of technology and contained Microsoft’s advances (.NET and it’s precursor DCOM) on it’s own Microsoft only island.

For the last fifteen years, Java EE has been  technology of choice for enterprises building scalable, reliable, sophisticated, typically business-critical applications and services. It’s proven to be one of the most important and long-lasting computing standards in our industry’s history – it’s hard (and a little painful) to imagine where we would be without it.

JBoss EAP (even before the Red Hat acquisition) was a market leader (as defined by industry analysts such as Gartner) and even today has continued its leadership position into new segments. It has lead many of the trends such as the original highly modular micro-kernel design (EAP 4 & 5), cloud-deployment, extensible module system, programmable management APIs, and first-class support for modern development tools like Maven and Git (EAP 6). That focus on bleeding-edge design has ensured that  JBoss EAP has always been ready to adapt to new architectures, deployment models and development practices. That is no less true today than it was ten years ago.

Today, IT organizations face many challenges – not least is the expectation that they can maintain large complex, intertwined application portfolios yet operate with the speed, efficiency and agility of web companies like Amazon, Netflix and Google. Unfortunately, few companies have the luxury of abandoning their existing applications to rewrite them as cloud-native micro-services using the latest functional, reactive framework.

I interact with enterprise customers on a regular basis and well understand the investment they’ve made in their Java EE applications and know well that those applications will be in active development and delivering value for years and decades to come. The owners of those applications – whether they are running on JBoss, Websphere or WebLogic need to know that their investment is safe yet they are able to adopt contemporary software delivery practices and new architectures.

Even as we achieve another major milestone in this product’s history (with the EAP 7 Beta) – we’re already looking ahead to understand how Enterprise Java will need to evolve to take advantages of the latest advances in cloud computing and enterprise architecture. But we do so without losing sight of the huge investment customers have made in Java EE to date. We’re pretty excited about the prospects of Java EE 8 but at the same time already experimenting with new ideas in projects like WildFly Swarm and vert.x. Customers want to embrace future advances in technology but have to be able to get the most from their application investment.

 

 

JBoss 2015 – What a Year !

Screen Shot 2015-12-18 at 1.36.25 PM

I’ve been part of the Middleware (aka JBoss) team at Red Hat for almost 8 years now and I can say pretty unequivocally that 2015 was a huge year. Huge. Huge in terms of growth (the team, revenue, customers); huge in terms of the number of new initiatives and markets we’re taking on and huge in terms of product releases. I don’t plan to enumerate all the year’s achievements here – there are way too many, but I did want to cover a few of the more recent announcements.

xPaaS

The first major milestone we hit with our xPaaS initiative – xPaaS being short-hand for taking our existing middleware products and capabilities and making them first class citizens of the cloud – specifically taking them to cloud environments via OpenShift (Red Hat’s PaaS). This is something we started back in 2012 with JBoss EAP and JBoss EWS on OpenShift 1.0 . In the meantime, OpenShift has been through a significant rearchitecture with 3.0 – adopting de-facto industry standard container technologies like Docker and Kubernetes. With the recent roll-out of OpenShift 3.1 – we’ve added support for additional middleware services :

  • Integration Services – supported JBoss Fuse – connects applications to the data sources, systems and APIs they need.
  • Real-time Decision Services – supported by JBoss BRMS for authoring, managing and executing sophisticated business rules.
  • In Memory Data Grid Service – supported by JBoss Data Grid – our market-leading,  distributed, in-memory key-value store that can boost your application performance and scalability by orders of magnitude.

That’s in addition to the existing Application, Web and Messaging Middleware Services already available on OpenShift. During our xPaaS journey – One of the major design goals for us has always been that no matter where you run JBoss – it’s the same – doesn’t matter if you’re running in a virtualized environment, directly on a public IaaS like EC2 or in our OPenShift PaaS – it’s still JBoss and behaves just the same. Developing apps. for EAP or Fuse on EC2 is no different to developing them for bare-metal or OpenShift. Read more about xPaaS here.

Mobile

Red Hat Mobile Application Platform (RHMAP) 3.6 – Red Hat’s MBaaS (Mobile back-end as a Service) was also released this week – while not a major feature release – this does demonstrate continued momentum (after the Feed Henry acquisition just over a year ago) and integration with existing Red Hat JBoss technologies- specifically the Unified Push Service which was in development upstream at the AeroGear project before we acquired FeedHenry. You can find more about RHMAP 3.6 here.

JBoss EAP 7

Finally we released JBoss EAP 7.0 BETA last week. This is a significant release in every respect; the last major EAP release (EAP 6.0) was back in June, 2012 and EAP 6 is now about 8 months away from entering its long-term maintenance phase. Among the major features in EAP 7 are :

  • New high-performance web subsystem based on Undertow – supporting Servlet 3.1, WebSockets, HTTP Upgrade
  • New high-performance messaging subsystem based on Apache ActiveMQ Artemis – same searing performance as HornetQ but expanded protocol support and Artemis is now the standard message broker for all JBoss products.
  • Supports Java EE 7 – full and web-profile and Java SE 8.
  • Enhanced (JSR-352) batch support – including cluster support, management of batch jobs and IDE (JBDS 9.0) integration
  • Improved upgrade experience from previous versions of EAP / WildFly and better support for competitive migrations using Windup.

I written have a longer blog post dedicated to EAP 7 if you want more detail.

As always – you can download the BETA via the Red Hat Customer Portal or from JBossDeveloper if you don’t have access. Release notes are here. And you’ll need some developer tooling to go with that – JBDS 9.0 is available for download here.

Here’s to a busy and productive 2016 – it’s going to be huge !

 

JBoss EAP 7 BETA Available !

Screen Shot 2015-12-19 at 2.25.41 PM

Today Red Hat released JBoss EAP 7.0 BETA.  This is the culmination of a couple of years of hard work by the JBoss EAP team at Red Hat and the broader WildFly community. I would like to thank them for their dedication and hard work and offer congratulations on achieving another huge milestone and major step forward in establishing Open Source in the enterprise and JBoss EAP as the open source standard for Java EE.

JBoss EAP 7  is a significant release in every respect; the last major EAP release (EAP 6.0) was back in June, 2012 and after 4 minor feature releases and numerous patch releases, EAP 6 is now just 8 months away from entering its long-term maintenance phase. While EAP 6 will be fully supported for many years to come – all new development and new features will target EAP 7 and beyond.

While Java EE 7 brings a major set of new features to EAP 7 (see below) it’s only a small part of what defines JBoss EAP 7. There are many other major updates in the release to help us keep up with demands of our customers, industry trends and align with other Red Hat products and initiatives.

  • New high-performance web subsystem based on Undertow – supporting Servlet 3.1, WebSockets, HTTP Upgrade
  • Undertow can also be deployed standalone as a lightweight, scalable proxy / load-balancer.
  • Move from JacORB to the standard OpenJDK ORB
  • New high-performance messaging subsystem based on Apache ActiveMQ Artemis – same searing performance as HornetQ with expanded protocol support and Artemis is now the standard message broker for all JBoss products.
  • Support for Java EE 7 – full and web-profile and Java SE 8.
  • Enhanced (JSR-352) batch support – including cluster support, management (liste, start, stop, resume) of batch jobs and IDE (JBDS 9.0) integration
  • Improved upgrade experience from previous versions of EAP / WildFly and better support for competitive migrations using Windup.
  • Improved JNDI, EJB, JMS and WS interoperability between EAP 7 and older versions – useful for side-by-side upgrades.
  • Ability to manage EAP 6 domain hosts and servers
  • Improved management console; easier navigation, and much better support for large scale domain configurations.
  • Graceful shutdown – allows servers to quiesce without aborting in-flight requests or transactions

Also the following features are available as Technical Preview :

  • A new JGroups based DistributedWorkManager
  • Execute JavaScript (using JDK8’s Nashorn), access JNDI and invoke CDI and JPA EntityBeans from JavaScript
  • HTTP 2.0 – connection multiplexing, header compression and server push

Major enhancements to Java EE 7 include :

A more detailed refresher on  on Java EE 7 features here.

As always – you can download the BETA via the Red Hat Customer Portal or from JBossDeveloper if you don’t have access. Release notes are here. And you’ll need some developer tooling to go with that – JBDS 9.0 is available for download here.

Give the BETA a try and give the team some feedback.

 

UP3 take 3

 Jawbone UP3 Band
 I pre-paid for the much anticipated UP3 band in December 2014, was dissapointed with the release delay and the downgrade from waterproof to splashproof but opted for the discount vs get my money back because I thought it was worth waiting for and worth waiting for Jawbone to get it right. I was dissapointed when my first band failed after just 6 weeks, more so  when the second band failed for exactly the same reason after just 3 months; as did my Wife’s first band after 3 months. After all this I decided enough was enough – I gave Jawbone a chance to explain how they have addressed the expanding case design issue. They didn’t even accept that it was a known issue so I demanded my money back – afer explaing the design flaw to a couple of tech support assistants and one manager I was told they wouldn’t be able to  give me a refund as my original purchase was beyond the 60 day limit and all they could do was send me a replacement. I couldn’t find that time limit easily in their returns policy.

So here I am after 6 months with my 3rd band and my wife with her 2nd band expecting both to fail before Ground Hog Day – essentially a reluctant customer. As most marketers know – if there’s one thing worse than a non customer  – it’s a reluctant customer.

Here’s the thing – the band (aside from previously mentioned design flaw) is very good – it does everything I want in the right form-factor. I think the iOS software is the best on the market in my opinion and they’ve done a decent job of iterating the band software to extend batery life and make sleep tracking easier. Their customer service rocks – they’ve never hesitated to process a return and done it very quickly each time – I suspect practice has made them good at this.

My message to  Jawbone – you can do better than this; you have to do beter than this – it’s a very competitive market. Redesign the band casing / battery, make it waterproof as you orginally claimed and give your loyal but possible reluctant customers and free / cheap upgrade path. 

Red Hat a leader in In Memory Data Grids (IMDGs)

Fast on the heels of the Forrester Report on Mobile Infrastructure Services where Red Hat’s Mobile Platform was positioned as a leader,  this week Forrester released a new report for In Memory Data Grids and again Red Hat is positioned as a leader.

As they say in the report “There Is No Better Way To Achieve Blazing Fast Performance At Scale” – something I’ve been saying since we released the first version of JBoss Data Grid a couple of years ago. Another way to think about this is that there’s no better way to ruin a perfectly decent user experience that having slow, un-responsive applications. In memory data grids offer a quick solution to this without the cost and risk of major application re-architecture.

As with all Red Hat products, Red Hat JBoss Data Grid is pure Open Source and is based on the upstream Infinispan project.

More information on Red Hat JBoss Data Grid here.

Red Hat a leader in Mobile Infrastructure Services

When we made the decision to acquire FeedHenry last fall (almost a year to the day) – we wanted to be sure we were getting behind the future generation of mobile platforms and not trying to compete in the past. FeedHenry’s cloud-native mobile platform is focussed on developing and deploying the back-end (wiring data-sources and services) while keeping the front-end app. development open to customer choice. It’s also a first for Red Hat for a product line supporting node.js as the runtime for developing application logic. JavaScript on the front-end and the back-end is pretty powerful and a mobile platform that doesn’t support that would be pretty limiting today.

More validation of our choice comes in the form of Forrester’s recent report – “Mobile Infrastructure
Services, Q3 2015″ – diagram shown below.

Forrester’s research uncovered a market in which AnyPresence, Kony, and Red Hat lead the pack. Appcelerator, Kinvey, IBM, Microsoft, Oracle, and SAP offer competitive options. MobileSmith is a market Challenger.

The Forrester Wave™: Mobile Infrastructure Services Q3 2015

Now that FeedHenry is part of Red Hat and the integration with Red Hat’s broader cloud and middleware portfolio is progressing that Red Hat Mobile Application Platform’s market presence will improve considerably and Red Hat will advance further up and to the right .

Red Hat has moved into this market through acquisition of FeedHenry, and is quickly moving to connect its backend-as-a-service offerings to Red Hat’s own extensive set of integration tooling, JBoss Fuse.

Congratulations to the Red Hat Mobile team – this is a great way to celebrate the first anniversary !

DevOpsDays DC 2015

Making the most of the calm before the storm that is Red Hat Summit, I raced up to Alexandria, VA for a few days to attend DevOpsDays DC 2015, hosted at the pretty stunning USPTO offices. I heard from the organizers they reached capacity 400+ and sold out within a few days.

IMG_2469

Interesting mix of neck beards and suits. And there were some very senior suits – eg. Mark Schwarz (CIO of USCIS) – he had a keynote but sat through pretty much *every* talk over the two days – he clearly thought that was a good investment. and I’m sure he’s a very busy guy. Very unscientific poll – 2/3rds of the people I spoke with were Ops (not Dev background) and about 50% were public sector (vs private).

The Public Sector companies / individuals who spoke at the show (or who I spoke with) were all bought into agile and DevOps (self selected group I guess) – huge possible gains for organizations working with / for the Gov. – huge challenges as well. On taht note, I recently came across this Wired article that provides some good insight into how Public Sector IT is changing.

Red Hat was probably the biggest company sponsoring – others were Ansible, Chef, Elastic, Puppet, Sonatype and a bunch of smaller companies (mostly public sector tech consulting)

Aside from Shawn Wells (Red Hat) co-preso. on OpenSCAP, the best talk was Joshua Corman (CTO at Sonatype) “Continuous Acceleration with a Software Supply Chain Approach”

tl;dr – open source usage is booming, becoming a higher priority target for hackers so more high severity CVEs than ever. Projects are often slow to react and release fixes, vendors are even slower and customers are even slower. Basically we’re all doing a bad job. Treat the software supply chain like Toyota do – fewer, better managed suppliers, higher velocity delivery pipeline (ie. DevOps). Ergo – use Maven to at least understand your dependencies.

Second best talk – Ken Johnson and Chris Gates “Devoops and how I hacked you”. tl;dr – don’t download and run random stuff from the internet unless you expect to get seriously pwned. Ran through popular OSS tools and outlined the most common exploits – pretty eye opening. You could have heard a pin drop if it weren’t for the noise of people txting their colleagues to check wether  X, Y and Z were patched and updated. Basic stuff – default , unencrypted passwords. Old versions with known, well advertised but fixed exploits. Adopt devops so it doesn’t have to take 6 months to roll out a new (secure) version of Jenkins, WordPress, Drupal, etc.

Schedule : http://devopsdaysdc2015.busyconf.com/schedule

Recordings : DevOps Day 2015 on Livestream

Breakouts :

I joined a couple of the Open Space breakouts but was more interested in seeing what was popular. Here are the top four :

  1. Docker overview – 50ppl – poll – who’s using in prod – just 2 – it was actually only 1 – the other was a pre launch startup
  2. CI / CD – 100% about Jenkins. Most of the questions about scaling and performance or resilience.
  3. Secure automation
  4. Burnout and suicide prevention (sadly)

All in a all a good value show – will definitely make the time next year – though I suspect they’ll need a bigger venue.