What do you like best about the product?

We've been using HAProxy for around a decade, initially starting with the Community Edition and gradually expanding its footprint. We used industry-standard SLBs from major vendors, but the exorbitant price tags and outrageous support costs were becoming unsustainable. Two years ago, we leaped HAProxy Enterprise Edition (HAPEE) with the Fusion Control Plane (FCP), and it's been a game-changer.

Here's what we love:

Performance: HAPEE's performance is impressive. Even without leveraging hardware acceleration (which they support through Xeon Platinum chips), it easily handles our traffic load.

Stability: Since migrating, we've experienced rock-solid stability. HAPEE has proven to be a reliable workhorse.

Full configurability: HAPEE offers complete control and customization, allowing us to tailor it precisely to our needs.

Fusion Control Plane (FCP): The FCP has revolutionized our workflow, enabling full transactional changes across our clusters. This was a huge pain point with our previous solutions.

Unbeatable Support: This is arguably HAPEE's strongest point. The support team is phenomenal. They're available through Slack, providing near-instantaneous responses and practical solutions. We've dealt with companies charging 10x or 30x more for support that pales in comparison. HAProxy's support team sets a new standard.

Cost-Effectiveness: The pricing for HAPEE is significantly lower than any other comparable solution on the market. It's a breath of fresh air after years of being squeezed by our previous vendors.

Community Edition: We also utilize the HAProxy Community Edition for simpler micro-deployments, and it also excels in that role.

We've moved to HAProxy EE, and the cost/performance/support are so good we are not looking back.

What do you dislike about the product?

Honestly, there's not much to dislike, but the Fusion Control Plane (FCP) has some growing pains. While it does many things exceptionally well, it's still under active development, and occasional kinks need to be ironed out, or some things you would expect haven't been implemented yet.

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

We use it at the edge and in internal clusters along with with keepalive to balance traffic and provide redundancy.