Welcome to The Long View—where we peruse the news of the week and strip it to the essentials. Let’s work out what really matters.
This week: Translating legacy COBOL code to a slightly more modern language, and Arm will go public (again) next month. (more…)
At its VMware Explore 2023 event, VMware today unveiled an orchestration tool, dubbed VMware Edge, that promises to simplify the management of edge computing services at scale.
At the same time, VMware has launched a Project Keswick initiative through which it will work with customers to define additional use cases for VMware Edge based on zero-trust principles and GitOps workflows for deploying applications. The company also unveiled a VMware Retail Edge platform based on VMware Edge and is planning to add additional offerings aimed at specific vertical industry segments.
Finally, VMware is making available a managed connectivity service through which it will provide private 4G/5G wireless services.
VMware Edge Cloud Orchestrator is based on a tool that VMware created to orchestrate the management of its secure access service edge (SASE) platform. That has now been extended to support a wider range of software running on edge computing platforms.
Sanjay Uppal, senior vice president and general manager of service provider and edge for VMware, said the goal is to streamline the management of software-defined edge computing platforms via a single console. Achieving that goal requires the ability to identify and programmatically manage workloads distributed across a mix of networks, he added.
VMware is betting that as more workloads are distributed to edge computing platforms, there will be a greater need to centralize the management of workloads running on platforms that are not always easy to physically access. Less clear, however, is to what degree those edge computing platforms will be managed by IT teams versus operational technology (OT) teams that are usually embedded within a department that manages a specific business process.
However, regardless of who manages those platforms, they will be relying more on GitOps workflows to programmatically provision infrastructure and deploy applications that will need to be regularly updated. In fact, because these platforms are often dispersed across hundreds of physical locations, the need for a pull-based orchestration tool that works asynchronously is critical, said Uppal.
In the longer term, IT teams should also expect artificial intelligence (AI) capabilities to be applied in a way that makes edge computing self-healing, he added.
No one can say exactly how many organizations have adopted GitOps to manage the provisioning of both infrastructure and the delivery of applications, but it’s conceivable that edge computing will force the issue. The tricky part will be finding a way to manage those GitOps workflows at the necessary level of scale. Most DevOps teams already struggle with implementing DevOps workflows at scale, so if more workloads are to be distributed across hundreds of edge computing platforms, there will be a much more acute need for automation to consistently deploy, secure and update applications.
Regardless of approach, however, there may come a day when the number of workloads deployed at the edge far outnumbers the number of workloads deployed in the cloud or in a data center. The challenge and the opportunity now is to get ahead of edge computing issues before they inevitably create a set of management issues that will be more difficult to address after applications have already been deployed.
Welcome to The Long View—where we peruse the news of the week and strip it to the essentials. Let’s work out what really matters.
This week: Rethinking return-to-office mandates and a ridiculous, ancient Windows bug. (more…)
Redis made a unified release of its namesake data store available today for consistent deployment across all supported platforms. Previously, the pace at which upgrades were made was staggered across different offerings only as they became available.
At the same time, Redis is moving to unify client libraries used to access its open source data store by working directly with the community maintainers of the five most popular client libraries—Jedis (Java), node-redis (NodeJS), redis-py (Python), NRedisStack (.Net), and Go-Redis (Go).
Other capabilities added to version 7.2 of the platform include support for auto-tiering, to more easily migrate less frequently accessed data to less expensive storage devices. This is managed via an enhanced interface for the Redis cluster management and data integration tools, available in preview. Both make it possible to consume data in real-time from another data source, such as a SQL database. Developers can also take advantage of triggers and functions to build those types of applications.
Redis is also previewing a search tool for vector data used to create generative artificial intelligence (AI) applications. Rather than employing a separate vector database to build these applications, the Redis platform supports this data type to enable generative AI applications to run alongside other classes of applications accessing the same data.
Redis CEO Rowan Trollope said collectively these efforts will help DevOps teams reduce costs and streamline the management of upgrades to the platform and updates to applications accessing it.
Most organizations initially adopt Redis to provide a layer of cache using a data store that runs in-memory. However, as more applications are developed many organizations wind up using Redis as their primary database for applications that require a lot of interactivity, noted Trollope.
While developers primarily make that architectural decision, the overall management of the Redis data store falls to DevOps teams that will now find it easier to streamline workflows across multiple instances of Redis running in the cloud, or on a Kubernetes cluster, added Trollope.
Of course, in some cases, organizations still rely on database administrators and data engineers to manage those workflows, but as DevOps continues to evolve, more instances of data stores are being incorporated into workflows managed by a DevOps team. The ultimate goal should be to align DevOps and DataOps best practices to increase the productivity of developers and data science teams building generative AI applications using vector data to customize an existing large language model (LLM).
Regardless of the motivation, what’s clear is the volume of data that needs to run in-memory continues to increase substantially. As that shift continues to occur, more latency-sensitive applications are being deployed. Maintaining the performance of those applications across a highly distributed enterprise environment with multiple service dependencies requires a level of DevOps expertise that is still difficult to find and retain.
Teleport has made available a chat interface, dubbed Teleport Assist, that makes use of OpenAI’s generative artificial intelligence (AI) application programming interfaces (APIs) to enable DevOps teams to troubleshoot infrastructure issues more easily.
Using the natural language interface (NLI) provided by Teleport Assist, DevOps engineers can ask questions and give instructions in English to massive fleets of servers. Capabilities include being able to suggest commands and scripts for debugging and resolving common issues and then generating commands for review before they are applied.
In future versions of Teleport Assist, support for other cloud resources, such as SQL databases and Kubernetes clusters, will be added. Teleport Assist currently supports Secure Shell (SSH) and bash scripts, with support for additional languages and interfaces to be added over time.
Teleport CEO Ev Kontsevoy said it’s already clear that many of the scripts that DevOps engineers rely on today for IT management will be replaced by generative AI platforms. The average IT administrator will be able to invoke DevOps platforms via an NLI that eliminates the need for extensive programming expertise, he added.
In effect, the NLI surfaced by generative AI platforms will serve to democratize DevOps expertise, Kontsevoy noted.
It’s still early days as far as the adoption of generative AI by DevOps engineers is concerned, but it’s already apparent that its impact will be profound. It will increase the pace at which applications are built while making it easier for software engineers to manage large codebases. Most of the generative AI advances thus far improve developer productivity, but tools such as Teleport Assist show how generative AI will also make it easier for more organizations to adopt DevOps best practices that previously were limited to organizations that could afford to hire and retain significant numbers of software engineers.
At their core, generative AI platforms are fundamentally changing the way humans interact with machines. Instead of requiring a developer or engineer to create a level of abstraction to communicate with a machine, it’s now possible for machines to understand the language humans use to communicate with each other. A a natural language interface will soon enable DevOps engineers to ask generative AI platforms to not only surface suggestions but also debug applications and even write code. The accuracy and quality of those results will vary depending on how the large language models (LLMs) that trained those platforms were, but over time there will be a wide range of LLMs that are trained to be used within specific domains such as DevOps.
The issue is determining how to organize DevOps teams as more tasks become automated. Generative AI doesn’t replace the need for IT professionals, but it does make it simpler to manage IT at unprecedented levels of scale. In fact, at this point, the proverbial generative AI genie is out of the bottle, so there is no going back. The only thing left to decide is how quickly generative AI becomes pervasively applied.
Mezmo has added additional capabilities to streamline the flow of telemetry data within DevOps workflows as part of an effort to make it simpler to surface insights and reduce the total cost of observability.
The company has also added integrations with more data sources to the Telemetry Pipeline platform, along with controls that make it simpler to optimize data storage and usage. For example, an Events-to-Metrics Processor capability identifies and extracts metrics from logs to make them easier for third-party tools to consume.
Mezmo CEO Tucker Callaway said these additions collectively make it simpler to surface insights that will make DevOps workflows more efficient. Previously, the company added capabilities such as rollback and redeploy, sequential parsing, error history management and data sample management to enable DevOps teams to manage telemetry data more efficiently.
In effect, Mezmo is making it easier to apply engineering best practices to the massive amounts of telemetry data generated across DevOps workflows, he noted.
No one knows for certain if all that data will require the addition of data engineers to DevOps teams, but as DevOps continues to evolve, it’s clear there is a need to cost-effectively manage data at scale. DevOps teams that embrace observability are quickly discovering that storing logs, traces and metrics for an extended amount of time can quickly become cost prohibitive. Naturally, there’s a lot more sensitivity to those costs during challenging economic times.
In theory, of course, various forms of artificial intelligence should automate data engineering best practices, but in the meantime, there simply isn’t enough data engineering expertise available. The need for a platform to optimize the flow of telemetry data across DevOps workflows is becoming more acute.
It’s not clear how quickly DevOps teams are embracing observability, but it’s clear that relying on predefined metrics to monitor IT environments is no longer sufficient. DevOps teams need to be able to query telemetry data to, for example, identify bottlenecks created by dependencies between microservices. Resolving those issues can’t be slowed down by a lack of data to query. Most problems can usually be traced back to a recent update to an application, but there are always going to be some issues that require access to historical data. That historical data should be quickly and easily available, especially since application downtime directly equates to lost revenue for many organizations.
Observability, regardless of cost, is rapidly becoming a requirement for DevOps teams’ success as application environments become more complex in the cloud-native era. Dependencies between the microservices that make up those applications are too difficult to track and analyze without the aid of an observability platform. The challenge is not every observability platform provider is equally concerned about how much it costs to store all that data. Striking the right balance between the amount of telemetry data collected and the cost of storing it has become a much larger concern.
Welcome to The Long View—where we peruse the news of the week and strip it to the essentials. Let’s work out what really matters.
This week: Google’s full-stack, browser based development environment in the cloud. It’s not just a Copilot clone, but aims to help you “get an app from zero to production.” (more…)
At its recent DASH 2023 conference, Datadog revealed Flex Logs and Intelligent Test Runner, two major new capabilities aimed at improving logging and testing efficiency for engineering teams.
Omri Sass, director of product management at Datadog, said Flex Logs introduces a new middle tier of log storage between real-time querying and cold archives. As Sass explained, this provides a cheaper option for retaining logs beyond the typical two weeks for occasional analytics and investigations. By decoupling storage from compute for reading logs, teams can independently scale each based on their needs. This enables retaining rich log data for months or years to uncover long-term trends, he said.
In addition, Sass noted that Flex Logs helps address common log analysis limitations around storage costs and rehydrating archives. The feature better aligns log pricing with specific use cases beyond just troubleshooting. According to Sass, this evolution in log storage tiers will likely shift more logs from legacy on-premises systems to its cloud platform.
The second capability, Intelligent Test Runner, leverages observability signals to optimize CI test suites. As Sass explained, comprehensive test automation in CI pipelines often increases deployment times and flakes. Intelligent Test Runner inspects code changes to determine associated tests, ignoring irrelevant ones, which helps avoid long wait times for minor changes just to exercise unaffected areas.
By tapping into existing telemetry around code commits, test health and runtimes, the feature adapts testing to each code modification. Sass explained that observability data has matured enough to enable real-time optimization like this. The end result is faster CI validation, so engineering teams can ship updates more rapidly.
Sass highlighted how Flex Logs and Intelligent Test Runner are Datadog’s attempts to address real pain points developers face around managing log data life cycles and CI speed bumps. Sass said he believes this combination of depth in core disciplines like logs and application performance monitoring (APM), along with expanding into adjacent developer workflows, is key to solving these issues through innovation.
Looking to the future, the observability space continues to evolve rapidly as emerging technologies and practices expand monitoring beyond traditional IT infrastructure. Companies handling modern workloads expect their observability platform to provide comprehensive visibility as development toolchains and architectures grow more complex, with the next generation of observability likely incorporating advanced analytics, automation and AI to help engineers tame systems at ever-increasing scale and speed.
Tricentis has extended its cloud service for testing applications to include the ability to emulate multiple classes of mobile computing devices.
Mav Turner, CTO for DevOps at Tricentis, said the Tricentis Device Cloud (TDC) eliminates the need for organizations to acquire multiple physical devices to test mobile applications at scale.
Capabilities include the tracking of key performance indicators (KPIs) such as blurriness, slow downloads, HTTP redirects and network or connectivity issues that adversely impact the user experience, in addition to performance issues such as load times and excessive consumption of CPU, memory and battery resources.
The grid-based service also uses machine learning algorithms to apply artificial intelligence (AI) to monitor application performance and detect video and audio quality issues.
Overall, the goal is to make it simpler for DevOps teams to build and deploy mobile applications faster and make sure they are thoroughly vetted rather than waiting for feedback from end users to discover issues that could have been easily identified before an application is deployed, said Turner.
The launch of TDC follows the company’s acquisition of Waldo, a provider of a software-as-a-service (SaaS) platform that provides IT teams with a no-code tool for generating mobile application tests. Waldo uses a virtual environment to simulate iOS and Android devices in a way that eliminates the need for physical devices to conduct tests. Instead, DevOps teams record their applications and then use the no-code tools to generate tests directly from within a browser.
In the meantime, the debate over how far application testing should be shifted left toward developers continues to rage. In theory, the more developers test applications themselves the fewer issues there should be once an application is deployed in a production environment. However, there is still going to be a need for dedicated testing teams to conduct tests across multiple stages of the application development process to ensure quality. In fact, as more routine tests are conducted by developers, it should free up application testers to focus more on edge use cases that might otherwise not be tested before being encountered by end users, noted Turner.
Testing mobile applications is more critical than ever because the cost of failure—in terms of impact on the business—is now much higher as more organizations depend on software to drive business processes. The challenge is that many of these app developers don’t have a software engineering mindset, so they need testing platforms that automatically create and run tests on their behalf in a way that can be customized for specific use cases. In time, many of those tests will be automatically created by artificial intelligence (AI) models that should reduce the cognitive load required to manage the testing process, added Turner. In effect, application testing will soon be democratized by AI, he noted.
Over time, the quality of applications should steadily improve as many of the low-level mistakes that developers make are eliminated. That should improve everything from application security to the user experience before mobile applications are initially deployed and then continuously updated.
At its DASH 2023 conference, Datadog today launched Bits, a digital assistant that uses generative artificial intelligence (AI) from OpenAI’s ChatGPT platform to provide real-time recommendations to resolve issues surfaced by the company’s observability platform.
At the same time, the company announced it is extending the reach of that platform to enable IT organizations to monitor the large language models (LLMs) many are now embedding within custom applications.
Datadog also released an LLM Observability platform in beta, which brings together data from applications, models and various integrations to help engineers discover and resolve AI issues in applications such as performance degradations, drift and hallucinations in real-time.
Those extensions complement an existing ability to monitor the usage patterns of application programming interfaces (APIs) making calls to the OpenAI ChatGPT platform.
Yrieix Garnier, vice president of product at Datadog, said the generative AI capabilities in beta are enabled by data the company used to extend the core capabilities of OpenAI’s ChatGPT platform.
Bits AI learns from both customer data and other sources to answer queries via a natural language interface that makes the Datadog platform more accessible to a much wider range of IT professionals, noted Garnier.
The overall goal is to reduce the mean-time-to-resolution for issues by, for example, explaining errors and suggesting code fixes that can be easily shared with developers via a Slack communication channel.
Those capabilities extend the functionality of the machine learning algorithms that Datadog has already embedded within its platform, noted Garnier.
While the generative AI capabilities added to the core Datadog platform will make DevOps teams more efficient, it’s not quite clear what role those teams will play in monitoring LLMs created by data science teams. In theory, DevOps teams should be responsible for optimizing the performance of any element of an application, but most DevOps engineers today don’t tend to have much experience with LLMs that are created by data science teams.
At the very least, however, DevOps engineers should be able to share the root cause of an issue that is adversely affecting those LLMs with those data science teams.
It’s not clear how quickly LLMs will be incorporated into applications, but most DevOps teams should expect them to soon be pervasive. In effect, Datadog is now providing DevOps teams with the AI tools they will need to effectively monitor applications infused with AI capabilities. The issue is not so much whether DevOps teams will be taking advantage of AI to better manage application environments but how quickly. Given the overall level of complexity DevOps teams regularly encounter, few are likely to want to take on that challenge without some level of AI augmentation.
The challenge is determining exactly what DevOps tasks might be best handled by a machine that enables software engineers to focus on more complex issues they previously didn’t have the time to address.