[Review/Test Request] Ubiquiti EFG / UXG-Enterprise

[galvesribeiro]

Hello folks!

I've just watch the "We used a $1M tool to test the HOT $279 Ubiquiti UCG-Fiber Gateway" video.

First of all, THANK YOU @Patrick and the team for being transparent and to be clear about the bad practices some manufacturers do forcing people to sign bizarre agreements in order to give them gear. Stay with an independent editorial and you will ALWAYS go further.

That being said, I saw on that video you testing that tiny home gateway (which is a great value product btw) and I wonder if you could run similar tests with "Enterprise" line of gateways, specifically the EFG and UXG-Enterprise.

The reason I'm asking this is because I have a 8 month old ticket with their support showing multiple evidences that those gateways can't route 25G traffic (or 50G aggregated) in ANY situation with same VLAN traffic getting close to 16G and cross-VLAN it can barely hit 1GbE. They don't acknowledge it publicly and keep pushing it as an "enterprise" product misleading people. I have they deployed on 8 sites that I own across office, home, and data centers, and all of them are behaving exactly same way.

It is not a hardware problem. The SoC on those are capable of easily handle 100G traffic.

I've provided to them thru the ticket everything needed to fix it but here we are, 8 months without having what we paid for.

It would be AWESOME if you could run the Cyperf tests against those devices and make a quick post or even a video. This would bring to light a pattern that we've seeing with Ubiquiti products.

"As always, thanks for the content and have an awesome day!"

 

[Patrick]

I know this is not exactly the answer you are looking for, but the plan now is to move up to higher end devices after we get the VPN and IDS/ DPI firewall workloads sorted.

Part of the reason is that it takes a lot of time configuring each unit, then the tool, so doing it twice is not ideal.

We are trying to operationalize this to make it easier to run many units through a standard test suite and do it once.

 

[galvesribeiro]

I know this is not exactly the answer you are looking for, but the plan now is to move up to higher end devices after we get the VPN and IDS/ DPI firewall workloads sorted.

Part of the reason is that it takes a lot of time configuring each unit, then the tool, so doing it twice is not ideal.

We are trying to operationalize this to make it easier to run many units through a standard test suite and do it once.

Got it. Thank you @Patrick.

I'm one of the people which fall for the Ubiquiti "Enterprise" marketing and had (now, wrongly so) bet on it expecting them to deliver while spending hundreds of thousands across multiple deployments.

Unfortunately there are no other content creators which are willing to say the truth about those devices because (guess what) as they are probably subject to the very same agreement you mentioned you refused to sign and they want to play the ball with Ubiquiti in order to keep getting free gear. It is a shame, so we have very few people with independent editorial like you guys.

The scenario I've got is that you don't even need that test gear you shown (although it would be great and fun ) but instead, simple tests like iPerf3 on the same VLAN and across VLANs connected directly on the EFG/UXG-E 25G ports and no switches would present the problem.

Ubiquiti is about to release new devices including a "EFG Core" which includes more 100G ports on it, but I'm afraid no matter what they do with the hardware, their software will be a crap bottleneck.

Nonetheless, thank you. Hopefully someone with public visibility will get to say the truth eventually so people don't fall on that mistakes I did.

Thanks!