You probably still want to know who actually built a given container and what’s running in it.
On overview of how Bloomberg is looking at the likes of Pivotal Container Services:
“Many Kubernetes distributions are good on day one, when they’re first deployed,” said Andrey Rybka, technical architect in the office of the CTO at Bloomberg, the global finance, media and tech company based in New York. “But what happens on day two, when something fails? Kubernetes doesn’t [automatically] address things like failures at the physical node level.”
The roadmap for Cloud Foundry Container Runtime includes support for stateful applications based on the StatefulSets feature that became available with Kubernetes 1.7 in June. The foundation also plans to integrate the Istio project, founded by IBM, Google and Lyft in May, which helps to manage network communications between microservices
While this is sort of a bummer story for Pivotal (we’d like to have this account), it has a good profile of American and their needs in it. All of which are representative of other large organizations, e.g.:
- Application types: “The first result is that the airline will migrate to the IBM Cloud some of its critical applications, including the main website, its customer-facing mobile app and its global network of check-in kiosks. Other workloads and tools, such as the company’s Cargo customer website, also will be moved to the IBM Cloud.”
- Managed data-centers/cloud: “The airline will be able to utilize the global footprint of IBM Cloud, which consists of more than 50 data centers in 17 countries, in addition to a wide range of application development capabilities.”
- Long-term planning: “We wanted to make sure that the cloud provider would be using Cloud Foundry and open-source technologies so we don’t get locked in by proprietary solutions,” Grubbs said. “We also wanted a partner that would offer us the agility to innovate at the organizational and process levels and have deep industry expertise with security at the core.”
- We want to do all the new meat-ware: “As part of this process, American will work with IBM Global Services to use IBM’s Garage Methodology of creating applications through a micro-services architecture, design thinking, agile methodology, DevOps and lean development, the company said.”
- Legacy, it’s how you got here: “IBM Cloud will help enable developers to build and change application functionalities for the airline’s customers. These customer-facing systems will be on the IBM Public Cloud, while American will maintain backend connectivity to other on-premise legacy and third-party systems, for true Hybrid Cloud functionality.”
- There’s a lot going on: “American Airlines and its subsidiary, American Eagle, offer an average of 6,700 flights per day to about 350 destinations in more than 50 countries. American has hubs in Charlotte, Chicago, Dallas/Fort Worth, Los Angeles, Miami, New York, Philadelphia, Phoenix and Washington, D.C.”
[O]ne well-publicized case in that vein, they said, was Home Depot directly working with Pivotal Software to introduce Pivotal Cloud Foundry to Google Cloud Platform. The home improvement retailer wanted to continue to use the popular development environment in the public cloud, but avoid giving business to Amazon’s largest profit-generating division.
A Pivotal spokesperson told CRN that Home Depot, like other Fortune 500 retail customers using Pivotal Cloud Foundry for app development, prefer Google Cloud Platform or Microsoft Azure above AWS. Pivotal and Google “rapidly accelerated joint R&D efforts to add new capabilities,” he said, “encouraged” by those retail giants.
At the same time, Pivotal and Microsoft have also stepped up efforts to integrate capabilities on Azure, “primarily driven by automakers,” he said.
New types of software and delivery mechanisms (SaaS, mobile) mean new problems and scale:
“We were so used to dealing with tens of servers and suddenly it was hundreds and thousands of servers,” which in turn created more work for the development teams.
“The digital expansion of business equals more work and firefighting,” Cox said.
Less time spent doing dumb-shit:
employees used to spend the eight hours of the park closed every night, manually updating each server. Now only one person can update the whole fleet in 30 minutes.
Some guiding principals and management challenges:
Cox said that leading a change of this order of magnitude involved three crucial ingredients:
1. Collaboration: break down silos, mutual objectives.
2. Curiosity: keep experimenting.
3. Courage: candor, challenge, no blaming or witch-hunting.
But these can come with its own leadership challenges, including:
• The politics of command and control.
• How new leadership can take a company in a new direction.
• The blame bias of who versus what.
And, some good motivation:
We keep moving forward, opening up new doors, doing more things because we’re curious.
There’s some good “how do I actually get my organization do all this unicorn stuff” comments in this interview with DreamWorks Animation’s Doug Sherman.
Here’s one sample bit on winning people over to microservices. Instead of going into the lab for six months to work on a tool that they think will be useful, they do a lot more user-driven work upfront and then do (it sounds like) weekly small batches to keep the users apprised of the tools and, you’d guess, give continuous feedback:
You have to understand what people want to do in their domain. In the past, Ive gotten it wrong. Ill come up with an idea I think is sound I think its the coolest thing ever and Ill work six months in isolation with my team, and then well do this big reveal. And every time we’ve done that, its gone horribly wrong, because 1) people feel like were lecturing to them, like we know better than them. And then 2) we would typically have over-engineered it! It would be like the 747 cockpit, you know? There would be this overwhelming amount of knobs and bits and pieces that I think are great to have, but from their viewpoint, they only need to do a few things, and thats an overwhelming amount of stuff to have to sign up to be able to do. So now, Ive gotten into a habit: before I even write a single line of code, I interview everybody that potentially will use the solution that Im going to write, and I keep them in lockstep with me and my team just about every week. We keep them engaged, helping to influence the direction Im basically trying to echo out in code all of what they want. Its gone so much better, because they feel invested. They don’t feel like in six months I’m revealing this big, mysterious thing. They feel like this is just something they’ve seen through iterations. And whats empowering about that, too, is if you can get the spiritual leaders of the different departments that you’re trying to encourage to use your solution, they’ll help sell it for you.
And then a bit on their progress:
Were about 50% of the way in having some amount of production coverage powered by microservices which are deployable in cloud containers powered by technologies such as Spring and Spring Cloud.
There’s more, good cultural change stories in the interview.
After all these years, print media still struggles versus the Internet. This long piece on how the travel magazine industry has been suffering covers many great topics. I suspect much of the analysis is the same for all of print media.
One of the problems is the new set of demands on writers in that field:
There is the pain point of figuring out an internal work flow that functions across platforms. Journalists, writers, and content creators often have specialized skillsets, so asking one to write a story, create a listicle, take photos, and film compelling videos about a trip is a major challenge.
“We just started working more efficiently that way and it really, it’s painful to integrate digital and print,” said Guzmán. “The plays are different, the workloads are different, the story ideation is different. In doing this, there’s this huge cultural shift that is exciting and difficult.”
And, then, even after suffering through all that “cultural shift,” the results are often disappointing:
“The iPad was just going to be this Jesus of magazines and I never really quite believed that because I knew how challenging it was was to rejigger the content to fit that format,” said Frank, who oversaw Travel + Leisure’s digital strategy in the early 2010s. “Having just gone through the process of signing up and downloading a magazine, it took forever and was buggy and it just wasn’t necessarily a great solution. I was never really bought the gospel that the tablet was going to be our savior. But we did it. I mean, we created a great app and it was beautiful. It won awards, but that was knowing what the usership was is a little disheartening.”
And, as ever, there’s the tense line between blaming “most reader are dumb” and “rivals are evil” when it comes to what’s to blame:
“I could have written the greatest travel story ever known, and it would not have gotten on the cover of the traffic oriented site because a Swedish bikini teen saved a kitten from a tree; which is going to be more popular?”
Let them watch cats.
Still, as the article opens up with, it’s the old Curse of Web 2.0 – former readers, now just travelers – writing the useful content in the form of reviews on TripAdvisor and such:
“In general, people don’t read a review and make a decision,” said Barbara Messing, chief marketing officer of TripAdvisor. “Consumers will read six to eight reviews. They might dig in a certain characteristic that they are interested in, maybe they really are interested in what the quality of the beach is, or maybe they are really interested in whether it’s kid friendly or not kid friendly. In general, people will hone in on the characteristics of something that’s most important to them, find that answer on TripAdvisor, get that most recent insights, check out the photos, check the forums, and really be able to make an informed decision of whether something is right for them. I think that the notion that people could rely on the wisdom of the crowd and the wisdom of individuals to their detriment, I just think that’s false, and I don’t think the reality is that is going to happen.”
There’s also some M&A history of trading various assets like Lonely Planet, Zagat, and Frommer’s back and forth as different management figures out what to do with them.
As ever, I’m no expert on the media industry. It seems like the core issue is that “the Internet” is so much more efficient at the Job to be Done for travel (as outlined by the TripAdvisor exec above) that the cost structure and business process from print magazines is not only inefficient, but unneeded. Those magazines are now over-serving (and thus, over-spending) with a worse product.
While the quality of TripAdvisor (and Yelp, for example) reviews is infinitely worse than glossy magazines, since there’s an infinite amount of more crappy reviews, with the occasional helpful ones…it sort of more than evens out in favor of Sweedish bikini cat rescuers. Plus, digital advertising has so much more spend (and overall, industry profit, if only by sheer volume if not margin) – it must be because it’s better at making the advertisers money and because it creates a larger market:
A brief note, from William Fellows at 451, on HSBC’s use of Google Cloud’s big data/analytical services:
They have lot of data, that’s only growing:
6PB in 2014, 77PB in 2015 and 93PB in 2016
What they use it for:
In addition to anti-money-laundering workloads (identification and reducing false positives), it is also migrating other machine-learning workloads to GCP, including finance liquidity reporting (six hours to six minutes), risk analytics (raise compute utilization from 10% to actual units consumed), risk reporting and valuation services (rapid provisioning of compute power instead of on-premises grid).
As I highlighted over the weekend, it seems like incumbent banks are doing pretty well wtih all this digital disruption stuff.
Thus far, it seems like the large banks are fending off digital disruption, perhaps embracing some of it on their own. The Economist takes a look:
- “Peer-to-peer lending, for instance, has grown rapidly, but still amounted to just $19bn on America’s biggest platforms and £3.8bn in Britain last year”
- “last year JPMorgan Chase spent over $9.5bn on technology, including $3bn on new initiatives”
- From a similar piece in the NY Times: “The consulting firm McKinsey estimated in a report last month that digital disruption could put $90 billion, or 25 percent of bank profits, at risk over the next three years as services become more automated and more tellers are replaced by chatbots.”
- But: “Much of this change, however, is now expected to come from the banks themselves as they absorb new ideas from the technology world and shrink their own operations, without necessarily losing significant numbers of customers to start-ups.”
- Back to The Economist piece: “As well as economies of scale, they enjoy the advantage of incumbency in a heavily regulated industry. Entrants have to apply for banking licences, hire compliance staff and so forth, the costs of which weigh more heavily on smaller firms.”
- Regulations and customer loyalty are less in China, resulting in more investment in new financial tech in Asia:
- As another article puts it: “China has four of the five most valuable financial technology start-ups in the world, according to CB Insights, with Ant Financial leading the way at $60 billion. And investments in financial technology rose 64 percent in China last year, while they were falling 29 percent in the United States, according to CB Insights.”
- Why? “The obvious reason that financial start-ups have not achieved the same level of growth in the United States is that most Americans already have access to a relatively functional set of financial products, unlike in Africa and China.”
- There’s some commentary on the speed of sharing blockchain updates can reduce multi-day bank transfers (and payments) to, I assume, minutes. Thus: ‘“Blockchain reduces the cost of trust,” says Mr Lubin of ConsenSys.’
Fixing legacy problems with new platforms, not easy
- The idea of building banking platforms to clean up the decades of legacy integration problems.
- Mainframes are a problem, as a Gartner report from last year puts it: “The challenge for many of today’s modernization projects is not simply a change in technology, but often a fundamental restructuring of application architectures and deployment models. Mainframe hardware and software architectures have defined the structure of applications built on this platform for the last 50 years. Tending toward large-scale, monolithic systems that are predominantly customized, they represent the ultimate in size, complexity, reliability and availability.”
- But, unless/until there’s a crisis, changes won’t be funded: “Banks need to be able to justify the cost and risk of any modernization project. This can be difficult in the face of a well-proven, time-tested portfolio that has represented the needs of the banking system for decades.”
- Sort of in the “but wasn’t that always the goal, but from that same article, Gartner suggests the vision for new fintech: ‘Gartner, Hype Cycle for Digital Banking Transformation, 2015, says, “To be truly digital, banks must pair an emphasis on customer-facing capabilities with investment in the technical, architectural, analytic and organizational foundations that enable participation in the financial services ecosystem.”’
- BCG has a prescriptive piece for setting the strategy for all this, from Nov. 2015.
- A bit correlation-y, but still useful, from that BCG piece: “While past performance is no guarantee of future results, and even though all the company’s results cannot be entirely attributed to BBVA’s digital transformation plan, so far many signs are encouraging. The number of BBVA’s digital customers increased by 68% from 2011 to 2014, reaching 8.4 million in mid-2014, of which 3.6 million were active mobile users. Because of the increasing use of digital channels and efforts to reconfigure the bank’s branch network—creating smaller branches that emphasize customer self-service and larger branches that provide higher levels of personalized advice through a remote cross-selling support system—BBVA achieved a reduction in costs of 8% in 2014, or €340 million, in the core business in Spain. Meanwhile, the bank’s net profits increased by 26% in 2014, reaching €2.6 billion.”
- And a more recent write-up of JPMC’s cloud-native programs, e.g.: ‘“We aren’t looking to decrease the amount of money the firm is spending on technology. We’re looking to change the mix between run-the-bank costs versus innovation investment,” he said. “We’ve got to continue to be really aggressive in reducing the run-the bank costs and do it in a very thoughtful way to maintain the existing technology base in the most efficient way possible.” …Dollars saved by using lower-cost cloud infrastructure and platforms will be reinvested in technology, he said.’ JPMC, of course, is a member of the Cloud Foundry Foundation which means, you know, they’re into that kind of thing.
Free food, during a limited, half-hour window, both saves people some hassle and gets them to show up at the same time to kick off the workday.
To understand why this is so important, picture Pivotal without free breakfast. Let’s start with the obvious. Most developers would sleep late if it were up to them. They’d roll into the office around 10 or 11 AM. Which means they’d grab a coffee, maybe respond to a few emails, and then sync up with the team.
Before you know it, the morning is over and it’s time for lunch. But hey, that’s okay, we live in a digital world, and you can show up whenever, so long as you get your work done, right? Wrong. Pair programming only works when you have people to pair with. And that means you need to sync their schedules.
We ring a cowbell at 9:05 AM. (The Toronto office smacks a golden gong with a mallet.) It signals that breakfast is over and the office-wide meeting is about to start. After the five-minute standup, the teams have their own standup meetings, and then pairs break off to get rolling at their workstations.
While posed as a pair programming enabler, take out pairing from the above and it also gets the point of having people show-up on-time, not dick around, and do actual work.
If you’ve seen me talk you know the joke of “how a developer spends their day” which usually includes 1-2 hours of actual coding because of all the meetings, you know, those 30 minute sitdown-standup meetings, architectrual reviews, deciding where to go to lunch, the post-lunch-buffet comma, “researching on the Internet, etc…. it’s all just unsynchronized schedules and little not attention spent on actually managing your staff’s time.