PerfectScale Pros and Cons

What I like best about PerfectScale is its ability to provide immediate, actionable insights into recommended CPU and memory requests and limits. By delivering real-time evidence and precise recommendations, PerfectScale eliminates the hours-long wait for service owners to assess issues, streamlining the resolution process and greatly enhancing day-to-day operations.

In our case, it helps us identify significant cost-saving opportunities, allowing us to reinvest in clusters that lack resources. This leads to a stable, resilient, and cost-effective environment without impacting our budget.

We use PerfectScale daily and have created Slack channels to receive alerts about services with gaps in their CPU and memory requests and limits. Currently, we are testing a new automation feature that adjusts JIT (Just-In-Time) CPU and memory requests and limits for pods of any deployment. As of writing this, it is working very well.

The implementation of their solution is very straightforward - just install their Helm chart, and the magic starts happening. If you want to use their new automation solution, you'll need to install an additional Helm chart, but it's not complex at all.

Their support and product managers are very responsive and open to feedback for improving UX and system features. Review collected by and hosted on G2.com.

In my opinion, their UI could be improved, and not everything is ready yet. There are still edge cases that their system does not cover, but they are working hard and quickly to address all gaps, bugs, and feature requests.

We are among the first adopters of their solution, and since we started using it, they have released many cool, new, and useful features. Review collected by and hosted on G2.com.

I've been using PerfectScale for a few months now, and I have to say, it's been a game-changer. Right from the start, the platform's user-friendly interface caught my attention. It's intuitive and doesn't require a steep learning curve, which is a huge plus when you're juggling multiple tasks.

The performance metrics and analytics are spot on, providing real-time insights that help us make informed decisions quickly.

What’s been a real bonus is how PerfectScale helps us identify waste in our processes and pinpoint where we can save on costs. It's not just about scaling up efficiently; it's also about doing so in a cost-effective way. The insights we've gained into our spending have made a noticeable difference to our bottom line, which is always a plus.

Another thing I appreciate is the support team. Whenever I’ve had questions or needed help, they’ve been prompt and genuinely helpful. It's clear they know their product inside out and care about their users' success.

In short, PerfectScale has made scaling our operations smoother, more efficient, and more cost-effective. If you're looking for a reliable platform that grows with your needs and helps you save on costs, I highly recommend giving it a try. Review collected by and hosted on G2.com.

Perfectscale agent pods consuming lot's of network bandwidth, which we have reported to perfectscale hoping to see improvement in future. Review collected by and hosted on G2.com.

Running Kubernetes clusters is no easy endeavour, and keeping costs under control while providing workloads with adeguate performance levels is a a rather time consuming task.

PerfectScale gives our engineers their time back to evolve our solutions, while the automation tunes our resource settings. This, combined with Karpenter, helps us to reduce the risk of under and over provisioning. Via the GUI it's easy to assess the effects of the tool and to visualise the metrics. Customer support is very proactive, and their guidance in achieving the best results has been outstanding. Our feedback has been taken into consideration, and new and useful features are added at a good pace. Review collected by and hosted on G2.com.

The tool is young but is rather solid. Some configuration options could be easier to configure, but I'm sure that given time even these minor quirks will be addressed. Review collected by and hosted on G2.com.

The level of details when it comes to examine your cluster:

TotalCost or CostperHour of a Pod (and of the specific containers), the Risk about your actual tuning of your workload (considering events like OOM and CPU Throttling) and the Waste/idle cost parameters (really useful to check if you are overprovisioning)

And for sure the Customer Suppport, they are always so helpful and caring! Review collected by and hosted on G2.com.

Maybe it could be not easy to manage at the beginning, considering all the features.

Especially on the Automation part: It's very useful but I would suggest first to check if there are some workloads that can have problem with her activation, so you can customize properly the configuration for exclude those in case Review collected by and hosted on G2.com.

What I really like about PerfectScale is how it seamlessly analyzes the workloads and makes intelligent adjustments without requiring constant manual intervention. Its ability to continuously monitor and adapt to changing demands in real time has made managing complex K8s environments much easier and more effective. Plus, the intuitive interface and insightful reporting provide clear visibility into resource utilization, helping to prevent over-provisioning and underutilization. Review collected by and hosted on G2.com.

the automation capabilities are currently restricted to deployments, leaving out statefulsets and jobs. This can be a drawback for those managing more complex workloads. Additionally, the product still has some early-stage issues, or “growing pains. Review collected by and hosted on G2.com.

This is the easiest to adopt and most usable Kubernetes workload optimization tool.

If you're focused on optimization, there's no better tool.

With a simple agent installation, it collects metrics from your Kubernetes workloads and recommends the right size.

The main and favorite features of this tool are...

- PodFit: recommend right-size of pods

- InfraFit: recommend right-size of nodes

- Automation: automatically rightsizes pods Review collected by and hosted on G2.com.

Sometimes webpages take a long time to render depending on the number and type of workloads.

This can be fixed with some configuration, but it requires work.

I'm sure this will be improved. Review collected by and hosted on G2.com.

The integration with our clusters and setup were easy and I liked how PerfectScale onboarded us and helped us every step of the way. We were able to take actionable steps quickly to bring down the cost of our Kubernetes clusters almost immediately. The recommendation engine for cpu and memory limits is robust and easily implemented. Finally, the autopilot feature has been working very well for us allowing us to save considerable cost with limited interaction and no negative performance impacts.

Finally, I like that PerfectScale treated us as a partner and took feature requests seriously and implemented them expodiciously. They have been a great company to work with. Review collected by and hosted on G2.com.

It is not necessarily a downside of using the platform but we are trying to tie pod level usage metrics to our cost of running our clusters and expose it through our cloud cost and usage reports. It would be nice if PerfectScale could help in this endeavor and we are working with them on a feature request to do just that. Review collected by and hosted on G2.com.

PerfectScale is a powerful tool for managing Kubernetes environments, offering several key benefits:

1. Automated Cost Optimization: PerfectScale uses AI algorithms to provide precise cost optimization recommendations, helping you reduce waste and save money

2. Enhanced Resilience and Availability: The platform autonomously right-sizes your environment, ensuring it is always scalable to meet demand. This enhances the resilience and availability of your applications

3. Real-Time Monitoring and Alerts: PerfectScale offers real-time notifications on resilience risks and cost anomalies, allowing you to address issues promptly

4. Seamless Integration: It integrates with various Kubernetes flavors and cloud providers, making it versatile and easy to use across different environments

5. User-Friendly Interface: The intuitive UI helps users quickly understand and act on optimization recommendations, regardless of their experience level Review collected by and hosted on G2.com.

More features on volumes, etc. will make it more useful. Review collected by and hosted on G2.com.

It is a powerfull solution that saves a lot of time to Ops teams with the question of sizing their containers.

When you're a big company or use large amount of containers this time saving become very important and valuable.

As a platform team, it is possble to deploy the solution and with no effort and then coach the platform users to use perfectscale. Having a SaaS solution with agent based is neutral in term on infrastructure consumption. A great match for those how use Datadog for exemple.

As a finops team, I can have a precise and already filtered view of what is happening inside clusters. It is something that is not available in the main public cloud providers.

It is a full story experience about sizing containers going from analyzing workloads to a set of recommandations and the final effective operations of automatically resize the containers.

The team is very responsive, feature requests process is in place. Review collected by and hosted on G2.com.

The solutions is still evolving in order to seemlessly add new clusters to the platform, good progress on that part but few features are still in the roadmap. When managing hundreds of cluster it could be challenging to manually active things in the UI.

The overall UI experience only manages clusters and worloads, the kubernetes Namespace is a very usefull level of agregation. It should be integrated in the solution.

We only use the recommandation mode right now because a lot of our internal customers are afraid of an automatic mode. It's hard to convince them even if we have insights from perfectscale platform. Recommandations are subjects to debates between experts, who's right in all the situations here is the question. Review collected by and hosted on G2.com.

Perfectscale.io is a wonderful tool that has emerged as one of our favourites at Client Projects - it not only finds wasted resources in compute, memory and areas where cost is an issue - it improves also performance. All the rightsizing and necessary steps are automated if chosen.

The other benefit is the cost and performance improvements with a more resilient deployment - anyone who runs expensive AI workloads in Container based deployments will appreciate this.

Also in our market we need specifically an easy and agnostic deployment method across different Container stacks - unlike other legacy tools that require a propietary and expensive to maintain Autoscaler this works with Karpenter, and other popular Autoscaling solutions which cost little now. It can be deployed on some European providers that other tools cannot do.

Support and the Engineering team are excellent via Slack and quick turnarounds on complex optimisation queries and it runs at an exceptionally low cost per vCPU hour. Really hard to beat in a market that looks for the best per Unit cost in expensive techstacks. Review collected by and hosted on G2.com.

I would like the UI to be improved a bit, in particular a light system toggle mode would be great -- although most deployments are done through the API and GitOps framework, its nice to switch into the UI on occasions and measure the Node consolidations and other items. Review collected by and hosted on G2.com.