Why There Is No Nutanix Review
-
Got the gist but that font makes me head hurt.
I am positive typewriters were never that bad. -
@momurda said in Why There Is No Nutanix Review:
Got the gist but that font makes me head hurt.
I am positive typewriters were never that bad.But Nutanix apparently is.
-
@jaredbusch said in Why There Is No Nutanix Review:
Well someone else has decided that Nutanix can piss off and posted a review..
http://www.cultofanarchy.org/perfomance-test-nutanix-4-node-cluste/
"We could do Linux or FreeBSD or some other real operating system instead of Windows"
HAH!
-
This is quite an entertaining read for sure.
-
Saved the article to my Wallabag install before it disappears into the ether.
-
So at the end of the article, they are saying data locality is really a horrible way to handle these kinds of systems. Which I think we all agree it is if you have a great network that can support the speeds your system is supporting.
@StarWind_Software to explain how StarWind vSAN can work around this issue.
-
@dustinb3403 said in Why There Is No Nutanix Review:
So at the end of the article, they are saying data locality is really a horrible way to handle these kinds of systems. Which I think we all agree it is if you have a great network that can support the speeds your system is supporting.
@StarWind_Software to explain how StarWind vSAN can work around this issue.
Overall the article is good, I think, and Nutanix definitely has major issues and always has - I mean it is downright famous for being a flaming pile of "can't pull the cart". Most hyperconverged solutions struve to be the horse pulling the cart, Nutanix has always been happy being the stuff that the horse leaves behind on the road.
But the issue here isn't data locality and shoud not be seen as that. Locality is really important and all the main vendors agree on this. Just because Nutanix does locality, and fails at other things, doesn't tell us that locality is bad, it just tells us that Nutanix screwed the pooch. It's physically impossible for distant data to compete with the speeds of local data, it is what it is. No matter how fast you make the interconnects, there will always be a benefit to having the main data reads be local.
-
@scottalanmiller the article states that even locally the performance was completely in the dumps. When the system was scaling it couldn't even reach the theoretical limits of what hardware as a whole should do.
-
"What’s interesting is however, pretty few companies maintain full local copies, most of the competitors prefer to chop the data into chunks and never keep single object as a whole."
This is the part that is really misleading. THis statement is true, but we have to understand the context. Let's take Scale, for example. Unlike Starwind who forces data to be local under "normal" situations, Scale just "encourages" it to be local. It's not that Scale doesn't do data locality, they just are more loose with it.
This is more an artefact of using RAIN instead of Network RAID. RAID, by definition, keeps full volumes together; RAIN has no such implication. RAIN is a broader topic, so not exactly apples to oranges. But with Starwind you could force all data to be non-local if you wanted, it's just that this is silly. With Scale, data might be non-local but will migrate to local for performance over time.
Part of the difference is that Starwind's approach focuses on massive speed requiriing a little less "pooling". Scale's approach focuses more on "single pooling" with less focus on tightly controlled performance.
THis means that with Starwind, for example, you will generally determine where your workloads run, not randomly migrate them around. But with Scale you will generally ignore where your workloads run and they will often move around on you.
-
@dustinb3403 said in Why There Is No Nutanix Review:
@scottalanmiller the article states that even locally the performance was completely in the dumps. When the system was scaling it couldn't even reach the theoretical limits of what hardware as a whole should do.
Yes, the performance of Nutanix, the issue is Nutanix, not locality.
-
@dustinb3403 said in Why There Is No Nutanix Review:
@scottalanmiller the article states that even locally the performance was completely in the dumps. When the system was scaling it couldn't even reach the theoretical limits of what hardware as a whole should do.
Imagine buying a Ferrari and taking it out in LA in heavy traffic. Then complaining that your Ferrari never goes over 15MPH. The issue is that they chose the wrong road (Nutanix) to test their Ferrari (NVMe) on, not that locality (cars) aren't faster than walking. They never really tested locality vs .non-locality, they never used a product that could showcase it. All of that stuff is background noise compared to the trainwreck that is Nutanix.
-
@nadnerb said in Why There Is No Nutanix Review:
Added to the list of dodgy cupcakes
New Tag: Dodgy Cupcakes
-
@black3dynamite said in Why There Is No Nutanix Review:
@nadnerb said in Why There Is No Nutanix Review:
Added to the list of dodgy cupcakes
New Tag: Dodgy Cupcakes
Very similar to sketch mojitos. If only we had that website.
-
Someone was discussing this today and I realized something that I had never noticed before. I have been told from someone with direct inside info that my
account was removed because I wouldn't tow the line and stop pointing out that Nutanix was doing stuff like this. Nutanix is a big advertiser there and had even tried to give me a trip to convince me to stop telling people the truth about how they operated. I declined.I never realized that the updating of this thread with the new information about them was just a day and a half before my account was removed over there - at a time when I had had no activity. This explains the timing so much more, Nutanix must have been watching this thread, seen that I had the link that they wanted to suppress, and acted to remove my account before the truth could get out over there.
It was the morning of February 28th when my account was removed, on a day that I was traveling so not even online and the activity on the 27th was essentially zero. I hadn't really been online for days. Which makes the timing of this so much more obvious now that I realize that they coincided.
-
@scottalanmiller said in Why There Is No Nutanix Review:
Someone was discussing this today and I realized something that I had never noticed before. I have been told from someone with direct inside info that my
account was removed because I wouldn't tow the line and stop pointing out that Nutanix was doing stuff like this. Nutanix is a big advertiser there and had even tried to give me a trip to convince me to stop telling people the truth about how they operated. I declined.I never realized that the updating of this thread with the new information about them was just a day and a half before my account was removed over there - at a time when I had had no activity. This explains the timing so much more, Nutanix must have been watching this thread, seen that I had the link that they wanted to suppress, and acted to remove my account before the truth could get out over there.
It was the morning of February 28th when my account was removed, on a day that I was traveling so not even online and the activity on the 27th was essentially zero. I hadn't really been online for days. Which makes the timing of this so much more obvious now that I realize that they coincided.
A buddy of mine who works for an insurance company, has been using Nutanix for a couple of years. He says that they like it. I do find it interesting how there is so much shady shit about them.
-
@wrx7m said in Why There Is No Nutanix Review:
A buddy of mine who works for an insurance company, has been using Nutanix for a couple of years.
Is he the administrator in charge of it?
@wrx7m said in Why There Is No Nutanix Review:
He says that they like it.
Who likes it, Nutanix or the business?
@wrx7m said in Why There Is No Nutanix Review:
I do find it interesting how there is so much shady shit about them.
You find it odd that Nutanix does everything in their power to hide their performance benchmarks against the competition and anything that is honest and makes their product look bad gets buried?
-
@wrx7m said in Why There Is No Nutanix Review:
A buddy of mine who works for an insurance company, has been using Nutanix for a couple of years. He says that they like it. I do find it interesting how there is so much shady shit about them.
The problem that we often see is that people evaluate hyperconvergence, not Nutanix. I bet if you asked him, he doesn't "like it" compared to the competition, just likes it in a vacuum, if that makes sense. I'm only guessing, but given that you are limited and slow compared to far cheaper options, what is there to like?
-
@DustinB3403 said in Why There Is No Nutanix Review:
@wrx7m said in Why There Is No Nutanix Review:
A buddy of mine who works for an insurance company, has been using Nutanix for a couple of years.
Is he the administrator in charge of it?
More importantly, is it his money being spent on it?
-
@wrx7m said in Why There Is No Nutanix Review:
I do find it interesting how there is so much shady shit about them.
Or you find it odd that there is literally so much shade on this brand, when no one else in this space is nearly as horrific about reviews of their product(s)?
-
@DustinB3403 said in Why There Is No Nutanix Review:
@wrx7m said in Why There Is No Nutanix Review:
A buddy of mine who works for an insurance company, has been using Nutanix for a couple of years.
Is he the administrator in charge of it?
@wrx7m said in Why There Is No Nutanix Review:
He says that they like it.
Who likes it, Nutanix or the business?
@wrx7m said in Why There Is No Nutanix Review:
I do find it interesting how there is so much shady shit about them.
You find it odd that Nutanix does everything in their power to hide their performance benchmarks against the competition and anything that is honest and makes their product look bad gets buried?
No, I don't think he deals with it directly. He is a Windows Admin and manages a ton of VMs that run on it. They are also using Cisco UCS, so who knows what TPTB know about the competition.
I think "they" refers to the team in their company that manages it.
I find it odd that they are so shitty, but are still around.
