Episode 152: Who put robots in my clouds? Oracle OpenWorld

Published: Oct. 24, 2018, 9 p.m.

There\u2019s all sorts of cloud stuff coming out of Oracle OpenWorld this week, so Brando and Cot\xe9 talk about the mouth-feel of the news. Related, Amazon\u2019s attempts to get off Oracle in Ohio, iCloud dropping out, and JEDI problems.

\n\nSponsored by SolarWinds\n\n

This episode is sponsored by SolarWinds and this week SolarWinds wants you to know about their DevOps tool: AppOptics.

\n\n

Today, there is a divide between application and infrastructure health metrics\u2014and the lack of unified dashboards, alerting, and management. With SolarWinds AppOptics you get a bird\u2019s-eye view across all your resources on a single pane of glass\u2014but can also drill quickly into the details.

\n\n

AppOptics includes built-in integrations for over 150 cloud-first applications, instant visibility into server and infrastructure performance, robust custom metrics dashboards, and automated APM request tracing. It\u2019s SaaS-hosted, easy to manage, and budget friendly.

\n\n

Over 275,000 customers trust SolarWinds for the performance data they need, and AppOptics lets developers and operations get back to doing what they love: delighting users.

\n\n

Learn more or try it free for 14 days, just go to appoptics.com/sdt

\n\nRelevant to your interests\n\n
    \n
  • Cloud, enterprise software to drive 2019 IT spending, says Gartner | ZDNet\n- CNCF Adopts Sysdig\u2019s Falco Container Runtime Monitor - The New Stack
  • \n
  • DTA goes cold on blockchain
  • \n
  • I think the DevOps people are into talking about \u201cproduct now\u201d\u2026?
  • \n
  • Ellison makes convincing pitch on automation and security for Oracle Cloud 2.0...but can\u2019t resist trashing AWS
  • \n
  • Kurt\u2019s summary of Oracle cloud stuff, pretty good.\n\n
      \n
    • Cot\xe9: Look, I don\u2019t really know their portfolio well. It\u2019s hard to follow cause it doesn\u2019t show up in all my feeds like, well, everything else. I\u2019m intrigued by their emphasis on performance and (to a lesser extent) cost. They really hit up the performance characteristics - I\u2019m not sure they mention ease of use or \u201coutcomes\u201d very much.
    • \n
    • The focus on security is bizarre. Not because they shouldn\u2019t have these things, but because these things are, well, what they should have. Cloud vendors don\u2019t go around chest thumping about how secure they are in the same way that bakers don\u2019t go around chest thumping about how their food is edible.
    • \n
    • Performance pitching has always been Oracle\u2019s thing (as those of us who used to read printed trade rags know). It\u2019s sort of indicative of easier sales: it\u2019s all numbers in a spreadsheet, then you sort a column and it tells you which vendor to pick.
    • \n
    • Then there\u2019s Oracle commentary on Amazon of how hard it is to move off Oracle, just barely wrapping itself in the mantle of \u201cbecause our stuff works better,\u201d when at the core it seems like the worst case of lock-in rent-seeking:
    • \n
    • \u2018Oracle Chairman and co-founder Larry Ellison isn't buying it. On the company's earnings call in December, Ellison said Amazon "is not moving off of Oracle." He reiterated his point at an August event, saying, "I don't think they can do it.
    • \n
    • \u2018They've had 10 years to get off Oracle, and they're still on Oracle," he said. "And it's not going to be easy for them to use their own technology. It's not going to be cost-effective. I mean, it's really, really hard.\u2019
    • \n
    • \u261e This kind of talk is why we all love to hear \u201cLarry\u201d (as everyone calls him) talk. He\u2019s like the Steve Banon of the IT industry.
    • \n
    • They should start demo\u2019ing at DevOpsDays and O\u2019Reilly conferences more.
    • \n
    • Topic: when pitching to \u201cthe community\u201d is irrelevant, or, \u201cCIOs don\u2019t go to your shit conferences, nerds.\u201d
    • \n
    • Now, to put me further out on on the ledge of not knowing Oracle well, they sell a shit-ton of ERP software. They could likely have a larger, positive impact on global productivity by making that ERP software better, no matter how good it is. In the coverage I\u2019ve read, there\u2019s little talk about how they\u2019re revolutionizing ERP stuff - how \u201cmachine learning\u201d is improving that. Can it figure out how to file expenses for me? Optimize a supply chain (what ever the fuck that means), etc.? For example, Oracle has the potential to turn all that Watson talk intro practical, everyday applications of \u201cAI.\u201d IBM doesn\u2019t have an ERP suite (they just have re-selling and packaging other people\u2019s stuff injected with Watson thingies - again, whatever the fuck that means) - but Oracle does, plus the foot-print of people using it. I\u2019m sure there\u2019s plenty of money in databases\u2026but their potential to improve their customer\u2019s life is probably more in apps.
    • \n
    • Diginomica had some ERP coverage: Park Hotels going from analog to digital in accounting, and an excellent example from Red Cross work on improving outcomes. And then back to our regularly scheduled price/performance talk.
    • \n
    • Kurt has some good, dry lines:
    • \n
    • Burn-town: \u201c[Oracle\u2019s] Cloud 2.0 looks more like Cloud 0.5\u201d compare to: \u201cMy project look like science fair, your project look like section 8.\u201d
    • \n
  • \n
  • Amazon's move off Oracle caused Prime Day outage in big Ohio warehouse, internal report says\n\n
      \n
    • \u201cThe outage, which lasted for hours on Prime Day, resulted in over 15,000 delayed packages and roughly $90,000 in wasted labor costs, according to the report. Those costs don't include all the lost hours spent by engineers troubleshooting and fixing the errors or any potential lost sales.\u201d
    • \n
    • I assume Amazon has saved much more than that by moving off Oracle.
    • \n
  • \n
  • Meanwhile, downtime effects us all: Apple iCloud down for (gasp!) hours!
  • \n
  • Topic: how much uptime do we really need? Cf. SRE last mile problems.
  • \n
  • US congress-critters question prime directive of Pentagon's $10bn JEDI cloud contract
  • \n
  • State of Wisconsin shares lessons learned on rolling out Oracle Exadata and how to reduce license costs
  • \n
  • HashiCorp updates its infrastructure automation suite for hybrid clouds
  • \n
  • An Alternative History of Silicon Valley Disruption
  • \n
  • Rockstar Games, crunch, and the great shame of the video games biz
  • \n
  • Apple\u2019s iCloud services suffered an extended outage
  • \n
  • Digital transformation of the week: \u201cEligible Travelers insurance customers will get a discount on their home insurance policies if they buy a smart home kit.\u201d
  • \n
  • Not everyone likes open spaces: \u201c7 is the magic number of team members for decision-making effectiveness. Once you reach that number, each additional member reduces effectiveness by 10%.\u201d
  • \n
  • Cloud Foundry Cult: \u201cThe users we spoke with didn't just see it as a PaaS \u2013 it was the underlying philosophy of application delivery and management upon which future developments would be based. The Foundation claims Cloud Foundry saves, on average, 10 weeks of development time and $100,000 per app development cycle. In fact, in its own survey, 92% of users cite cross-platform flexibility as important. If these panelists are gaining such benefits, it's easy to understand why they are so enamored with it.\u201d
  • \n
  • 300 VMs per admin is the magic number:\n\n
      \n
    • \u201cPrivate clouds owned and self-managed by enterprises can be cheaper than public cloud. The magic number to beat is about $25 per VM-month at 100% utilization. If the cost of the whole stack comes in under this number, then even with the addition of labor to manage that private cloud, it should be cheaper than public cloud. Obviously, with better labor efficiency, unit costs versus public cloud are lowered further, and the relative value of benefits increases. Enterprises unable to achieve a labor efficiency of 300 VMs per engineer are unlikely to beat public cloud on price.
    • \n
    • \u201dPartially managed clouds have good economics. If an enterprise is able to manage just the datacenter element of a private cloud at a ratio of at least 400 VMs per engineer, that cloud may cost less to operate than fully managed alternatives. We believe enterprises could easily beat this ratio.\u201d
    • \n
  • \n
  • Related: \u201cOf that, private cloud spending [on hardware] reached $4.6 billion, an increase of 28.2 percent year over year. That's a significant increase, but not as great as the jump in spending on public cloud IT infrastructure, which was $10.9 billion, a 58.9 percent year-over-year growth.\u201d
  • \n
\n\nConferences, et. al.\n\n\n\nListener Feedback\n\n
    \n
  • John from Australia wrote in to tell us he bought a T-Shirt and now needs stickers.
  • \n
\n\nSDT news & hype\n\n\n\nRecommendations\n\n
    \n
  • Brandon: Making a Murder Season 2.
  • \n
  • Cot\xe9: Staying in the same hotel when you go to a city. Consider the Lobster. Anti-recommendation: Logitech Slim case from iPad Pro with keyboard. The Apple one with a pen holder is probably better?
  • \n

Sponsored By: