IBM Turbonomic SaaS

What do you like best about the product?

As a Software Engineer who collaborates with both the Development and the SRE teams, I find Turbonomic to be a valuable tool as it grants visibility and strategic insight to our hybrid cloud infrastructure, which is composed of both AWS and on-prem VMware. Our architecture is complex, as we focus on building internal enterprise applications such as our Human Resource Management System (HRMS), as well as our fully automated billing system. The architecture also utilizes microservices along with Node.JS and Python.

Before Turbonomic, it was mostly guesswork when managing resources for the environments we provisioned. We had to provision more CPU and Memory than required to work around latency issues during peak times. Turbonomic’s solution offers a considerable amount of information pertaining to the resources and how they impact the applications. The solution has also aided managing a lot of things related to Kubernetes pods. It doesn’t show us a single hot node and call it a day. It shows us the various automations it can carry out to ‘remove’ some resource burden by allocating pods and nodes dynamically.

It also addresses the gaps between engineering and operations. If we have to deploy a resource-intensive billing system, the operations team can examine the dependency mapping to understand how the new system impacts the overall infrastructure and what trade-offs can be made in the systems to accommodate the new billing system.

What do you dislike about the product?

We can expect this to take considerable time to set up as a connector to all your tools, spanning APM, cloud service providers, Kubernetes, vCenter, etc.

In addition, gathering the many alerts and recommendations produced by the system will take considerable time. The system will become very pushy and nagging if the recommendations are not acted on. For example, one of the recommendations was to downscale a certain number of workers to optimize the cloud resources. One of the workers was used to process background tasks. The recommendation was based on the worker being underutilized on average, but it was actually downscaled the worker that processed the tasks. If time is not spent continually to manage the system, then the system will have to be configured to operate without any automation.

The user interface is probably the biggest hurdle for the non-technical members of the team, and probably the most overwhelming facet of the system, namely that it is not user friendly and gives little indication of your overall system performance versus your cloud costs.

What problems is the product solving and how is that benefiting you?

The principal problem that Turbonomic addresses for us is eliminating the waste of resources while consistently meeting performance thresholds. In an enterprise environment, the default engineering thinking when there is a problem is to just throw more hardware at the issue. With Turbonomic, we have objective data to more accurately determine the correct size for AWS instances and the container resources we need.

Our monthly cloud billing has gone down noticeably due to the more accurate sizing. The manual monitoring burden on our DevOps engineers has significantly decreased. In the past, our engineers would need to spend hours looking at the Grafana dashboards for resource bottlenecks. With Turbonomic, the DevOps engineers can be reallocated to more productive activities like application development. The burden of resource monitoring has been taken off of our engineers and placed on Turbonomic, which is a more accurate monitoring solution. Our engineers can focus on the business priorities and core applications instead of dealing with resource allocation and planning activities with much less engagement with the infrastructure planning teams.

What do you like best about the product?

I like IBM Turbonomic for its intelligent automation that automatically rightsizes resources to maintain performance while reducing cloud costs. It continuously analyzes application demand and adjusts CPU, memory, and cloud instance sizes in real time. This helps prevent performance bottlenecks, eliminates over-resourcing, and maintains SLAs.

What do you dislike about the product?

It can be complex to set up, has a learning curve, and licensing can be expensive for larger environments. Setup has been complex due to integration with multiple cloud platforms, hypervisors, and permission configurations. Initial setup is quite challenging because it required configuring integrations, permissions, and policies across environments which took some time to get right.

What problems is the product solving and how is that benefiting you?

IBM Turbonomic automatically optimizes application performance and infrastructure costs, solving issues like over-provisioning and cloud cost waste by rightsizing resources across environments.

What do you like best about the product?

I like that Turbonomic goes beyond basic monitoring by providing automated resource optimization. It helps ensure applications get the right amount of CPU and memory, while also avoiding over-provisioning. Its Kubernetes and cloud integrations make it especially useful in modern DevOps environments.

What do you dislike about the product?

The learning curve can feel steep at first, and making sense of all the optimization actions takes some familiarity with infrastructure and Kubernetes. For beginners, the interface and underlying concepts may come across as complex and a bit overwhelming initially.

What problems is the product solving and how is that benefiting you?

IBM Turbonomic helps address over-provisioned and under-utilized infrastructure by continuously analyzing application demand and automatically optimizing how resources are allocated. Rather than manually tweaking CPU, memory, or scaling policies, the platform recommends or can execute actions that keep performance steady while cutting unnecessary cloud and infrastructure costs. For me, this means less time spent on manual tuning and more confidence that my applications are running efficiently without overspending.

What do you like best about the product?

I really appreciate how IBM Turbonomic reduces the bill for our clients without them needing to get involved in all the technical details. It's pretty neat how this tool has helped us build trust with our clients, enabling them to scale their companies with more tools from us. In a sort of indirect way, this has boosted growth for both our business and our clients' businesses.

What do you dislike about the product?

So I think the biggest hurdle with it is its steep learning curve and initial configuration complexity which can delay the time-to-value for clients when they always expect instant results. While the core of the platform, which is AI, is really awesome, but the reporting and dashboarding thing is really rigid, making it difficult for me to present some high-level business-centric KPIs that I have to mostly present and is asked for. So I need a lot of manual data work to do while implementing and generating KPIs. Similarly, the licensing cost is really high and tough for those small-scale Salesforce implementations that makes me unable to pitch this tool for them.

What problems is the product solving and how is that benefiting you?

IBM Turbonomic's AI-driven autopilot optimizes resource allocation in real-time, balancing performance and cost-efficiency. It eliminates over-provisioning guesswork, reduces client cloud bills, and maintains high application speed, fostering client trust and business growth while ending 'war room' culture.

What do you like best about the product?

The AI-driven automation that continuously optimizes resource allocation across our hybrid cloud environment. We used to multiple iterations of performance testing to find the optimised value of CPU, Memory threesholds for our HPA which can be avoided now by leveraging IBM Turbomic in production.

What do you dislike about the product?

Steep learning curve and requires significant configuration time to get started. The interface could be more intuitive for new users. For Eg - Datadogs or ArgoCD dashboards are more intutive.

What problems is the product solving and how is that benefiting you?

For every application or service that we want to deploy, we use HPA for the scaling part in production. We have to do multiple iterations of performance testing to find the threeshold values of CPU and memory utilisation and min, max replica to provide to devops which they can configure in the HPA for the application to scale in production. This is a very resource and cost intensive process as we have to simulate the production traffic scenario multiple times this is replaced by IBM Turbonomic as it directly predicts these values based on previous utilization metrics of other applications and only 1-2 iterations are sufficient to come to the optimum configuarion.