a reckoning for tech by the humans that work with it

Hardware is the new Software – TechReckoning Dispatch v3n5

The TechReckoning Dispatch, Vol. 3, No.. 5. Sunday, March 27,  2016.

Hardware is the new software

Software is eating the world” is the famous quote from Marc Andreessen, a man that my generation of techies treated both as our exemplar and our measure by which we are found wanting. He’s our Mark Zuckerberg. (He Did What By What Age?) He’s now King of the Silicon Valley Venture Capitalists, which is good, because he pretty much always figured he was the smartest guy in the room.

The lessons of the last thirty years are that software wins over hardware, and that standard x86 hardware wins over anything else. But laws change. Moore’s law is choice, not an actual law. Twenty years ago we would have said that Windows wins over everything. Not so much anymore. So let’s posit that software and the Internet are changing everything. But how is the actual business of software doing?

Software, it turns out, is actually pretty hard to sell. Companies that use software – doing great. Open source software – the new normal. Commercial software margins, however, are trending relentlessly down. Think about your own experience – do you like paying Oracle and VMware and Microsoft for the honest fruits of their labors? Or do you always have the nagging feeling you’re getting ripped off? I’ve been reading The Software Paradox by Redmonk cofounder Stephen O’Grady, which I highly recommend. (Stephen also was the author of The New Kingmakers, another win. He’s got a way with the booklet. You bloggers should try it some time.)

So if nobody wants to pay for software anymore, what’s a poor company to do? If you look at a whole group of recent infrastructure solutions, they got people to pay by bundling in standard  hardware. Nutanix, SimpliVity, Scale, Coho Data, Rubrik, even EMC — all are bundling x86 servers with their software, in part so they can charge you $100K without you grumbling. Leveraging the Dell x86 OEM capabilities were no doubt one of the drivers of the EMC transaction. The margins on x86 hardware aren’t great, but it makes the top line revenue so much more delicious, and we’ve gotta buy hardware from somebody, so why not you?

Another thread to untangle, however, is the VMware hardware dampening effect. VMware grabbed much of the value of the datacenter for the last 10 years. Other infrastructure components, especially storage, needed to work well and operate in standard ways with VMware to retain traction. But now we’ve worked out those platform interfaces and we’re starting to see investments in really interesting new hardware coming to market:

Throughout my career, people have been trying to sell me non-standard hardware to eke out more performance, and the smart answer was generally no thanks, Is the future still a fabric of white box x86 servers or are we entering into an age of value-added hardware?

Worth A Click

The Google Cloud Platform Next 2016 conference was this week. The consensus seems to be the the conference gave a good showing, a more customer- and enterprise-oriented event. However, enterprise folks wonder about the focus on new-fangled Googly cloud features — with much of the world still vaguely worried that they might have some Windows Server 2003 boxes around, touting machine learning and serverless computing in your cloud does not warm a cold CIO heart.

Google Cloud — Impressions from GCPNext keynote by @CloudOpinion at Medium. Short, pretty good overview. Worth a click if that’s all the time you have.

Ex-Netflix cloud guru and Battery VC dude Adrian Cockcroft is the biggest evangelist for the AWS way of life I know, so he’s got a pretty sober assessment. Cloud at Scale – Is Google Ready for Enterprise Yet?

Urs Hoetzle talked about what comes next for Google: For one thing, he said “NoOps” application development will get easier with automation from App Engine and Kubernetes. He thinks we need automated security at scale based on immune systems for security deeply embedded in the apps and networks. He made the point that server-less architecture is deeply embedded in GCP. Use of the term “NoOps” is likely to turn off enterprise customers, who have large operations teams.

If Diane Greene can pull off this transformation, people might actually spell her name right. She is getting good marks for some pretty basic stuff, though, as the following quote highlights. She actually went out and visited customers and partners, which was evidently a pretty rare thing at Google. Everybody who has seen an enterprise vendor executive travel schedule will be chuckling to themselves about now. Google’s Greene Hastens Cloud Expansion to Catch Amazon by @jackclarkSF at Bloomberg.

Greene is also changing the way Google sells and markets. She’s hiring across the board and demanding staff work more closely together and talk to customers more often. That includes creating a team that meets with enterprise customers to ensure Google is building what they need. … This year, Greene came to a crucial customer meeting with SADA Systems, a big Google partner. It was the first time in nine years an executive of her seniority attended a customer meeting, said SADA Chief Executive Officer Tony Safoian.

(Note that everybody’s coming at AWS with their MBA smarty pants on. Microsoft is of course leveraging their whole stack. Google is including Google Apps under Greene’s purview for their salvo. IBM, Oracle, Salesforce are all coming with applications. The IaaS Wars are over. Begun, the Cloud App Wars have.)

A Greene profile from Cade Metz at Wired. The Tech Exec Who Wants the Cloud to Be Google’s Moneymaker

At VMware, the engineers worked very closely with our customers, closely with the field, and it was exciting for everyone. So, one of things I did here at Google when I arrived was to combine sales, marketing, engineering, and product. This is powerful. Engineers love having an impact on what customers can do, and by bringing (the Google cloud staff) together in this way, we can work towards that, feeding off each other and moving more quickly.

For a broad summary, read 10 big announcements from Google’s Cloud Conference by Brandon Butler at Network World/CIO.

If there is one big takeaway from GCP NEXT, it is Google’s major focus on machine learning. Google introduced a product family this week with multiple machine learning tools, a cloud-based platform for building predictive analytics models based on data customers have stored in Google’s cloud. “For example, a financial services app that predicts values using regression models, or a classification service for images,” Google explained on its blog announcing the service. “Cloud Machine Learning will take care of everything from data ingestion through to prediction. The result: now any application can take advantage of the same deep learning techniques that power many of Google’s services.”

Keep Up With Your Neighbors

Upcoming: theCUBE #BigDataSV event in San Jose this Weds.  Silicon Valley VMware User Group on April 12 in San Jose. ZertoCON in Boston on May 23-25. I also do weddings and bar mitzvahs.

A great interview with VMware’s William Lam on the latest Geek Whisperers. In case you missed it, I also produced the Mayfield Chat with Champions podcast — check out Season 1 – interviews with the CEOs of Docker, Nutanix, and more.

We are still looking for a part-time community manager at TechReckoning. Know someone that wants to pick up 10-20 hours each week working remotely and who is humane and detail-oriented and who likes geeks? Let’s talk.

Always Read The Comments

Slack really is a phenomenon, and our commenters come at it from very thoughtful angles. I tried to connect Slack and Bots and Echo, but perhaps we’ll tease them apart later.

Sean Duffy:

I enjoyed the discussion around Slack this week. I’m a member of three different teams and find it is great to catch up on the overall state of affairs in a team, however I always ensure that I disable notifications on my phone and in browser so that I don’t get too distracted. (This is actually my general policy for all chat/realtime collaboration apps!)

On the subject of ChatOps, I’ve developed a small NodeJS chat bot that hooks up to Slack’s API and sends web requests off to various REST endpoints to pull information from different systems I use at work. I’ve given the bot a list of commands that you can issue it in a chat room (appears as another bot user, similar to slackbot), and it will respond back with the information requested by spitting out stats, figures, or even charts using Google’s charting API. It started out as a fun project, but is now easily extensible and serves to make life a little easier for certain queries.

Sam McGeown:

I’m a member of several slack teams…I’ve found that the usefulness of it varies with the people on it, and the number of people in channels. The “water cooler chat” aspect of slack is fun for a little while, but it becomes a nightmare when you’re trying to work and some people are having a giphy war (don’t get me started on giphy). Of course you can mute the notifications for a channel and try and ignore it, but that kind of defeats the object of a communication tool doesn’t it? For me there seems to be a sweet spot of about 30 people in a channel – any more than that and the noise outweighs the usefulness of the channel. On bigger slack teams that means being involved in more focussed channels and muting the general social chit-chat. Don’t get me wrong, I think slack is a fantastic tool that provides a great way for distributed teams to keep in touch, but as with all tools, it’s effectiveness is in how it’s used.

Eric Wright:

As an attempted Slack user, I really dug in.  It didn’t stick.  I wrote a little ditty about it here: http://discoposse.com/2016/03/21/all-slacked-out-long-live-email/

In a world where collaboration is king, I think that it just didn’t stick the landing for the teams that I work with.  And despite many people taking the Apple antenna approach on my story by telling me that I’m holding it wrong, I am not alone in having trouble finding the sweet spot with it.

Adding Slack to the overall arsenal of collaboration could work.  I’m sure that the same was said for Orkut.  We all know how that worked out.

Rob Nelson:

Slack, hipchat, chat ops, all violate the number one rule of operations: if it’s not in the ticket, IT NEVER HAPPENED! Yes, you certainly CAN connect them to a ticket system, but do you now want to maintain two systems for the same purpose, and increase the chance of critical information not making it into the ticket log?

This isn’t really restricted to operations, either. Developers likely have Jira tickets instead of Remedy tickets. Tasks and work complete are captured somewhere in an authoritative system that persists across time.

This isn’t to say these various IM (and soon phone and video) tools have no place. IRC is great to communicate across companies and groups. Company IM tools are great to communicate about the issue at hand (esp with 3+ players). And they’re all great for real time discussions, say when a tech gets on site and is replacing gear or during code review or when sales is asking shipping when a customer’s parts have been on the loading dock for three days.

But they don’t replace ticket systems, so you must use care in how they are used. Like everything else, use them in moderation.

Michael Stump:

I personally like the concept of Slack. I’m part of a Baltimore tech-based Slack that gets only ~80 people online a day. But it’s been months since I even launched the app. Perhaps if my 9-5 coworkers were Slack types (and not the grumpy, crumpled life-long launchers of .bat files and MMC snap-ins) I’d like it more. And maybe it’s a sign of my age, but I’ll take in-person collaboration over these things any day of the week. And since my team is all local, a tool like Slack seems like a good solution to a problem that we don’t exactly have.

Chris Wahl:

I enjoy the concept behind Slack and, like any communication platform, think that you get out of the system what you put in. Using filtering of notifications and leaving irrelevant channels is important, just like tagging and setting up filters for email is important. Surprisingly, one of the most annoying things with Slack is the archaic need to sign into different teams using unique credentials on a thick client – I’d rather just log in once (especially on a new device) and have my entire set of Slack teams load with my full preferences. It’s when folks try to use Slack for task assignment that I tune out, because there’s few ways to really flag something for follow-up and create actionable assignments without heavy lifting in IFTTT or something similar. I already have this power in e-mail, and it’s already configured for task management (with an integration to Todoist) so I don’t bother trying to re-create that in Slack. Much like Twitter, I feel no obligation to read everything that was said in Slack. If my attention is needed, I’ll wait for a mention or a @here / @channel.

As for ChatOps – I’ve found it helpful for the small team that I work with. I use a few integration points into Slack to control and monitor a lab environment. Creating the integration appeals to my need to automate things, and I find that minimal effort can be expended to gain control over 80% of what I do on a regular basis. That’s handy, especially since others can see what I’m doing and learn by example. My bot is a bit sassy and bitter, akin to Marvin the Paranoid Android, because that’s how all bots should be. :)

Thomas LaRock sums it all up:

Slack has made some things better, but at the end of the day it is still text-based communication and that mode of communication is painfully inefficient to actual conversations. That being said, I’d rather be on Slack than in a meeting. The real trick, as with any tool that promises to enhance productivity, is to find out what works well when compared to other tools and to use it for that, and not to think it will replace everything.

Just Hit Reply

Is hardware the new hotness? Use your favorite hardware and send an email to jtroyer@techreckoniong.com and I’ll include your comments next week.

(And my last request: if you think somebody might find this interesting, please forward (and if this was forwarded, please subscribe. Like Rocky, we aren’t going to quit until they give us an Oscar.)

The TechReckoning Dispatch. A periodic newsletter from John Mark Troyer with links and opinions about enterprise technology. I like people, but I also like not being around people. Archive. Subscribe. Email. “And that Lorax?… He didn’t show up any more. But the next week he knocked on my new office door. He snapped, I’m the Lorax who speaks for the trees which you seem to be chopping as fast as you please. But I’m also in charge of the Brown Bar-ba-loots who played in the shade in their Bar-ba-loot suits and happily lived, eating Truffula Fruits. NOW…thanks to your hacking my trees to the ground, there’s not enough Truffula Fruit to go ’round. And my poor Bar-ba-loots are all getting the crummies because they have gas, and no food, in their tummies! They loved living here. But I can’t let them stay. They’ll have to find food. And I hope that they may. Good luck, boys, he cried. And he sent them away. I, the Once-ler, felt sad as I watched them all go. BUT… business is business! And business must grow regardless of crummies in tummies, you know.”
Copyright © 2016 TechReckoning, All rights reserved.