High School Management Method: Blame a Customer
June 9, 2021
I noted another allegedly true anecdote. If the information is correct, gentle reader, we have another example of the high school science club management method. Think acne, no date for the prom, and a weird laugh type of science club. Before you get too excited, yes, I was a member of my high school’s science club and I think an officer as well as a proponent of the HSSC approach to social interaction. Proud am I.
“Fastly Claims a Single Customer Responsible for Widespread Internet Outage” asserts:
The company is now claiming the issue stemmed from a bug and one customer’s configuration change. “We experienced a global outage due to an undiscovered software bug that surfaced on June 8 when it was triggered by a valid customer configuration change,” Nick Rockwell, the company’s SVP of engineering and infrastructure wrote in a blog post last night. “This outage was broad and severe, and we’re truly sorry for the impact to our customers and everyone who relies on them.”
Yep, a customer using the Fastly cloud service.
Two observations:
- Unnoticed flaws will be found and noticed, maybe exploited. Fragility and vulnerability are engineered in.
- Customer service is likely to subject the individual to an inbound call loop. Take that, you valued customer.
And what about Amazon’s bulletproof, super redundant, fail over whiz bang system. Oh, it failed for users.
Yep, high school science club thinking says, “We did not do it.” Yeah.
Stephen E Arnold, June 9, 2021
Reconciling Two Views of Cloud Computing
June 2, 2021
I think it would be helpful to read “The Cost of Cloud, a Trillion Dollar Paradox.” The write up is an MBA team effort, and it makes what I think is an interesting point. The cloud makes sense when a company is small and doing the “go fast, break things” stuff. But as the company becomes larger, the cloud becomes expensive and slaps handcuffs on the customer. The MBAs may not agree with my précis, but it works okay for me.
Then read “Atlassian Claims It’s a Step Closer to Achieving Nirvana with Its Data.” The main point of the essay to that centralizing cloud work is better, faster, and all around more wonderful than the multi-cloud thing. I winced at the use of the word “nirvana.” Amazon AWS and nirvana don’t fit together like peanut butter and chocolate or pinga and salt. (Nirvana, I think, means according to Google’s recycling of the Oxford “languages”
transcendent state in which there is neither suffering, desire, nor sense of self, and the subject is released from the effects of karma and the cycle of death and rebirth.
That’s AWS for sure.
Both articles are marketing material. The a16z piece makes it clear that the firm’s analysts are on the ball. I think the message is, “We’re on the ball. We put money where it will really pay off.” The Ziff story is a marketing tchotchke, and it is designed to send a specific message about the freedom from suffering, desire, etc. associated with the use of AWS services.
What’s the nitty gritty?
Marketing, not analysis nor personal experience, has become the payload of what appears to be technical relevant information. This is a good thing, right. Perfect for home economics and political science majors who wrangle jobs in or around technology.
Lock in and cost control are not difficult concepts in my opinion. Pick one. Nirvana is near.
Stephen E Arnold, June 2, 2021
What Is Cloud Computing? It May Be Timesharing REbranded
June 1, 2021
I have been around long enough to watch hot trends come and go. Then years or decades later the “old” new thing returns. “Nvidia Is Renting Out Its A.I. Superpod Platform for $90K a Month” states:
Nvidia is looking to make work and development in artificial intelligence more accessible, giving researchers an easy way to access its DGX 2 supercomputer. The company announced that it will launch a subscription service for its DGX Superpod as an affordable way to gain entry into the world of supercomputers.
Does this sound like timesharing to you? It does to me. And what about those automatic upticks in charges? It is too early to tell, but my hunch is that there will be “peak times,” data transfer thresholds, and a taxi meter method applied to some user actions. I hope I am wrong, but, hey, timesharing business models have been around since — what? — the 1950s. That is long enough for those thrilling moments after opening a timesharing invoice to become one of the benefits of this “new” but “old” approach to computing.
Will the Nvidia supercomputing deals include a white coat? One tip: If you tour the superpod data facility, take a sweater.
Stephen E Arnold, June 1, 2021
A Field of Data Silos: No Problem
May 5, 2021
The hype about silos has followed data to the cloud. IT Brief grumbles, “How Cloud Silos Are Holding Organisations Back.” Although the brief write-up acknowledges that silos can be desirable, it issues the familiar call to unify the data therein. PureStorage CTO Mark Jobbins writes:
“Overcoming the challenges presented by having cloud silos requires organisations to develop a robust data architecture. Having a common data platform should form the foundation of the data architecture, one that decouples applications and their data from their underlying infrastructure, preventing organizations from being locked into a single delivery model. Working with a multi-cloud architecture is valuable because it helps organizations utilize best-in-breed services from the various cloud service providers. It also reduces vendor lock-in, improves redundancy, and lets businesses choose the ideal features they need for their operations. It’s important to have a strong multi-cloud strategy to ensure the business gets the right mix of security, performance, and cost. The strategy should include the tools and technologies that consolidate cloud resources into a single, cohesive interface for managing cloud infrastructure. Hybrid clouds bring public and private clouds together.”
Such “hybrid clouds” allow an organization to retain those advantages of that multi-cloud architecture with the blessed unified platform. Of course, this is no simple task, so we are told one must recruit a gifted storage specialist to help. We presume this is where Jobbins’ company comes in.
Cynthia Murrell, May 5, 2021
Alphabet: Another PR Hit Related to Raising Prices and Changing the Google Rules?
April 23, 2021
Here in Harrod’s Creek, everyone — and I mean everyone, including my phat, phaux phrench bulldog — loves Google. After reading “Why I Distrust Google Cloud More Than AWS or Azure” it is quite clear that the post in iAsylum.net is authored by someone who would find our Harrod’s Creek perception off base.
The write up contains some salty language. On the other hand, there are a number of links to information supportive of the argument that Google cannot be trusted. Now trust, like ethics, is a slippery fish. In fact, I am not sure my trust checkbook has much value today.
The main point of the iAsylum write up is that Alphabet Google cannot be trusted. The principal reasons are that Google changes prices and acts in capricious ways. Examples range from Google Map fees to the GOOG’s approach to developers.
The most painful point for us lovers of all things Google was the question in the essay:
Will Google Cloud even exist a decade from now?
That’s a difficult question to answer. Some companies are predictable. Amazon’s Bezos bulldozer moves in quite specific directions. True, it can swerve to avoid a large rock, but for the most part, the Bezos bulldozer’s actions are not much of a surprise. Got a hot product? Amazon may just happen to have one too. No surprises.
Google is unpredictable. There’s the HR and ethics mess in the AI unit. There’s the spate of legal challenges about the firm’s approach to advertising. There’s the search service which returns some darned interesting results, often not related to the query the user submitted.
For those of us in Harrod’s Creek, worries about the future should be factored into our lives. But for now, we love those Google mouse pads. Our last remaining mouse pad is now yellowed and cracking. But it once was a spiffy thing.
Let me rephrase the iAsylum question:
Will Google Cloud evolve like my Google mouse pad?
Stephen E Arnold, April 23, 2021
Oracle Matches One Amazon AWS Capability: Bringing Order to Chaos
April 19, 2021
In 2018, I started noticing more Amazon AWS support for ServiceNow. ServiceNow is a company which uses cloud technology to help its customers manage digital workflows for enterprise operations. Amazon revealed in 2018 “How to install and configure the AWS Service Management Connector for ServiceNow,” the procedure which some AWS customers had mastered before the blog post gave its stamp of approval.
“Oracle Integrates ServiceNow into its Cloud Infrastructure” makes it clear that the much loved database vendor is doing what AWS did in 2018. The article reports:
Oracle has announced the integration of ServiceNow into its Oracle Cloud Infrastructure. The integration means enterprise customers have the ability to access and manage OCI (Oracle Cloud Infrastructure) resources via their existing ServiceNow service portal and the ITOM (ServiceNow IT Operations Management) Visibility application, which will give them a single dashboard to manage their public cloud resources from Oracle and other cloud providers.
Legacy Oracle customers like government agencies are likely to find the integration helpful. At one time, the likes of Amazon itself and Google might have been over the moon. Both of these cloud giants jettisoned Oracle technology and have moved in other directions.
A ServiceNow VP spins the Oracle move this way:
“With this integration, ServiceNow and Oracle are making it seamless for enterprises to unlock productivity for distributed teams to deliver products and services faster, access powerful business insights and create great experiences for employees, wherever they may be,” says ServiceNow’s vice president & general manager of Operations Management & Data Foundations, Jeff Hausman. Joint customers leveraging the Now Platform and OCI will get the best of both worlds, a seamless experience that maximizes the value of cloud investments and the ability to harness the power of artificial intelligence for proactive operations.”
Many buzzwords like seamless, unlock productivity, business insights, experiences which are “great”, value, proactive, and of course artificial intelligence.
The winner may be ServiceNow. For Oracle, I am not sure yet. Maybe on deck to enter the cross cloud de-chaosizing work now going on in many organizations.
Stephen E Arnold, April 19, 2021
Amazon AWS EC2 Pricing
February 11, 2021
Amazon AWS makes many things simple: Off the shelf machine learning models, buying cables, and spending money. If you want to get a sense for the complexity of pricing at AWS, take a look at “EC2 Instances.Info: Easy Amazon ED2 Instance Comparison.” The effort required to compile the table was significant. In addition to the data structured by EC service, region, and other tags — there’s the splash page table itself. Impressive. For those with some financial and technical expertise, a new job category now exists: Figuring out AWS pricing for a project and then determining how to minimize costs over time. From the Amazon one click patent to this pricing inventory. How far has Amazon driven the Bezos bulldozer? A long way.
Stephen E Arnold, February 11, 2021
Can a Cockroach Love the Google Cloud? Absolutely
February 9, 2021
Cockroach Labs has released its third annual report comparing cloud service providers Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP). On its own blog the company posts, “GCP Outpaces Azure, AWS in the 2021 Cloud Report.” The focus is on online transaction processing (OLTP). Writers Arul Ajmani, John Kendall, Yevgeniy Miretskiy, and Jessica Edwards tell us:
“Our intention is to help our customers and any builder of OLTP applications understand the performance tradeoffs present within each cloud and within each cloud’s individual machines. Perhaps your current configuration isn’t the most cost effective. Or you are looking to build a net-new application and want to see which provider has the fastest network latency. Maybe storage has been an issue in the past and you are looking for new solutions. Regardless of your motivation, the report is designed to help you achieve your goals and develop the best architecture for your specific needs. The 2021 Cloud Report is developed by a team of dedicated engineers and industry experts at Cockroach Labs. It compares AWS, Azure, and GCP on micro and industry benchmarks that reflect critical OLTP applications and workloads. This year, we assessed 54 machines and conducted nearly 1,000 benchmark runs to measure CPU Performance (CoreMark), Network Performance (Netperf), Storage I/O Performance (FIO), OLTP Performance (Cockroach Labs Derivative of TPC-C).”
The post summarizes the report’s highlights. As suggested by the title, the team found Google to deliver the most throughput. On the other hand, AWS’ network latencies remain on top for the third year in a row. We’re told AWS’ custom Graviton2 Processor beat the competition, both running AMD processors, for 16-core CPU performance. The writers also explain when it is worth paying more for each providers’ “advanced disks.” For more details, see the post or navigate to the report itself. Cloud SQL database maker Cockroach Labs was founded in 2015 and is based in New York City. No observations about the prevalence of certain insects in Alphabet City.
Cynthia Murrell, February 9, 2021
Subscriptions Are Dead: Bad News for Substack and Its Truck Load of Competitors
February 3, 2021
I know. I know. I know that “Subscription-Based Pricing Is Dead: Smart SaaS Companies Are Shifting to Usage-Based Models” is talking about cloud service providers. These are the small, emotionally sensitive firms like Amazon, Google, Microsoft, and others who struggle to make ends meet each month. The basic idea is that the taxi meter approach to pricing is the future. Hop in the cab, tell the head in the clouds driver your destination, and pay what the meter shows upon arrival. Did your driver crash? Did your driver take you to Sonic Drive In before reversing course and delivering you near your destination? Did your driver like some gig workers driving vehicles for money pull a gun and rob you? No? Lucky you.
The write up states:
Some fear that investors will hate usage-based pricing because customers aren’t locked into a subscription. But, investors actually see it as a sign that customers are seeing value from a product and there’s no shelf-ware. In fact, investors are increasingly rewarding usage-based companies in the market. Usage-based companies are trading at a 50% revenue multiple premium over their peers. Investors especially love how the usage-based pricing model pairs with the land-and-expand business model. And of the IPOs over the last three years, seven of the nine that had the best net dollar retention all have a usage-based model.
To read this article, guess what? You have to pay a subscription fee. I know. I know. Silicon Valley “real” news outfits just emit parental and oracular, consult like statements.
A couple of observations may be warranted:
First, many customers dislike usage based pricing because of surprises when the bill is presented. And, believe me, when the bill is submitted, getting a sensitive firm to alter it can be a time sink hole.
Second, the usage based model was one that was popular among some timesharing companies. Example: The much loved Dialcom or the European Space Agency’s operation decades ago. Why? Surprise fees.
Third, usage based pricing demands convoluted price lists. I assume that you, gentle reader, remember the wonderful days of IBM’s J1, J2, and J3 fee schedules. AT&T had some excellent methods as well. After Judge Green’s break up of Ma Bell, even Baby Bells howled when Bellcore fired off an invoice. Those were the days.
Now, if the write up is correct, the good old days have returned, except at the “real” news outfit making this profound statement.
Stephen E Arnold, February 3, 2021
AWS Offers Multicloud Services Without Fanfare
January 21, 2021
One problem with cloud offerings is when a service does not sold for one operating system, but not another. AWS usually brags about its accomplishments, but Protocol said, “AWS Quietly Enters The Multicloud Era.”
AWS has two new versions of its managed containers and managed Kubernetes services, EKS Anywhere and ECS Anywhere, that can run on both Microsoft Azure and Google Cloud. AWS confirmed that its new software will manage workloads running on other cloud providers. AWS does not like to play with other cloud services, however, its customers do, so they caved.
Google and Microsoft were late to the multicloud game too. When their customers demanded software management on multiple cloud infrastructures, they realized many used AWS. It was also a good way to make their customers happy and possibly more money.
AWS lacks support, though:
“It does not appear that ECS Anywhere and EKS Anywhere offer the same degree of support for multicloud deployments as Azure Arc or Google Anthos, which were designed to be user-friendly multicloud tools. And according to the ECS Anywhere launch blog post, ‘the supportability of ECS Anywhere scenarios at the time of general availability may be artificially limited due to other constraints.’”
That stinks for the moment, but give it awhile and the AWS team will offer more support as demand for service grows.
Whitney Grace, January 21, 2021
-
- Subscribe to Beyond Search
Feature archive
News archive
-
