What's new

Fear and Loathing of The Cloud

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

thiggins

Mr. Easy
Staff member
Discussion moved

These posts originally were in this thread. But since the discussion is broader than eero's product, I have moved them here.
 
It seems that all user control is done via the cloud, which means that if they go out of business, or decide to kill the servers, or drop support, then your product will become an overpriced plastic brick.
 
Last edited:
Cloud based 'anything'. All that's wrong with manufacturers today.

Meaning they cannot deliver a working, stable and reliable example of what they promise gullible consumers and with a price that would make both happy to do business with each other.
 
It seems that all user control is done via the cloud, which means that if they go out of business, or decide to kill the servers, or drop support, then your produce will become an overpriced plastic brick.
I avoid such proprietary products. Beyond dependance upon their servers, if that small startup fails, and you are dependent on their proprietary networking, then when you need a replacement or addition, you are SOL.

That's why we need the IEEE to promulgate a mesh standard. Years go by and there is none in consumer-space - due to lack of need.

The serious meshes are from Tropos and Cisco, with backhaul on a proprietary mesh and user access on a different radio (usually) with vanilla 802.11/WiFi.
 
It seems that all user control is done via the cloud, which means that if they go out of business, or decide to kill the servers, or drop support, then your produce will become an overpriced plastic brick.
True, but Meraki wasn't going anywhere, and nor is their system as it's been purchased by Cisco.
Cloud based 'anything'. All that's wrong with manufacturers today.
I don't think that's the case. It's hard to manage more than a single router without having various login screens, syslogs and whatnot to get a good picture of a large network. Most enterprise stuff has moved to cloud management a long time ago, aggregating the entire network in one place that it can be manged. It's not really 'the cloud' in the way some pie in the sky implementations make it sound, but more like a hosted integrated remote management console that's free and part of the product (that's what we have with Meraki anyways).

Now, a lot of the enterprise consoles aren't free. And if you don't buy the license fee or support fee, your $10k+ router becomes a paperweight. I have a serious problem with that.
I avoid such proprietary products. Beyond dependance upon their servers, if that small startup fails, and you are dependent on their proprietary networking, then when you need a replacement or addition, you are SOL.

That's why we need the IEEE to promulgate a mesh standard. Years go by and there is none in consumer-space - due to lack of need.

The serious meshes are from Tropos and Cisco, with backhaul on a proprietary mesh and user access on a different radio (usually) with vanilla 802.11/WiFi.
I don't think there will ever be an IEEE mesh standard. Forget the home market, there's not enough demand in the corporate and business market. Everyone has a type of mesh implementation (like the Cisco and Tropos you've mentioned), but it's tied to that vendor. And you are right--if they go kaput, so does your network. But as long as its cost-effective while it's up, who cares? Network equipment is getting upgraded these days before it fails anyways. Just look at how crappy computers are made these days. When I built a 486dx-33 back in the mid 1990s, it never even had a slight issue running even 10 years later. You can't say that today about almost any computer that's not an expensive server.
 
I don't think that's the case. It's hard to manage more than a single router without having various login screens, syslogs and whatnot to get a good picture of a large network. Most enterprise stuff has moved to cloud management a long time ago, aggregating the entire network in one place that it can be manged. It's not really 'the cloud' in the way some pie in the sky implementations make it sound, but more like a hosted integrated remote management console that's free and part of the product (that's what we have with Meraki anyways).


That version of 'cloud' that Meraki uses I might be able to live with.

I was talking about the true 'cloud' that happens on servers / equipment other than what I bought. That implementation is simply $tupid.

I don't buy hardware that is crippled with such viruses like reliance on cloud control and at the mercy and whims of corporations. And my customers would not understand why their equipment didn't work just because the original company went out of business, or was sold 3 to 6 months ago either.

Sure, huge savings can be made for the big corporations that do take a chance on these options. But they also have the money to jump on another train if / when disaster strikes their first bet. Not so easy for smaller companies to eat these kinds of losses. For many, it may even be the end of their business, depending on the specifics of how and what was affected by a major part of their network / security / backup infrastructure simply dissolving in front of their eyes.
 
That version of 'cloud' that Meraki uses I might be able to live with.

I was talking about the true 'cloud' that happens on servers / equipment other than what I bought. That implementation is simply $tupid.

I don't buy hardware that is crippled with such viruses like reliance on cloud control and at the mercy and whims of corporations. And my customers would not understand why their equipment didn't work just because the original company went out of business, or was sold 3 to 6 months ago either.

Sure, huge savings can be made for the big corporations that do take a chance on these options. But they also have the money to jump on another train if / when disaster strikes their first bet. Not so easy for smaller companies to eat these kinds of losses. For many, it may even be the end of their business, depending on the specifics of how and what was affected by a major part of their network / security / backup infrastructure simply dissolving in front of their eyes.
I'm right there with you. I was not liking the idea of Meraki's management console because of the exact same possible fate. Luckily, their APs will resort to their last known configuration if they lose touch with the cloud or it ever goes down permanently. And since my install was pretty much permanent and nothing was moving, I was okay with that.

I think the reasons you cited above is why a lot of smb equipment doesn't rely on the 'pushed' or 'managed' services that enterprise providers require. IT can't be this huge continuous cost that goes on forever, especially if it can just be taken away because of something beyond the business's control.

This is exactly why I opted to use our used (and sometimes glitchy) vpn routers that we had lying around to build our site-to-site networks. Sure, I could have bought Meraki's products and had a one-click cloud managed vpn, but we would have been paying on that forever. We needed something to be one and done.
 
One and done. The key to a successful business model.

At least one with satisfied customers. Even though everything seems to be going the other way right now.

One that nickels and dimes it's customers to death may be great for the shareholders, but they will eventually fold from their own weight.

A pyramid scheme comes to mind right about here...
 
One and done. The key to a successful business model.

At least one with satisfied customers. Even though everything seems to be going the other way right now.

One that nickels and dimes it's customers to death may be great for the shareholders, but they will eventually fold from their own weight.

A pyramid scheme comes to mind right about here...
And it's the stupid consumer mindset that has ruined everything. Back in the 1950s, the consumer mindset was wanting something 'built to last'. Now, it's 'I want the latest and greatest that will last until the next upgrade'. That's fine and dandy for expensive spoiled teenagers and their cell phones, but that doesn't fly for companies that have to lay a technological foundation and then use that to get real work done.

I remember back in the 1990s when there was no such thing as having to upgrade the desktop every 5 years. Then the whole 'upgrade every 3 years' was the going motto in the corporate world until the economy crashed. Then it came back to tco (total cost of ownership) and things seem to be on a more relaxed upgrade path with a frantic 'patch and update' routine.

And networking equipment was very 'set it and forget it' back then. You set up the configuration and the device would just WORK. Now, with buggy firmware and the like, it's the desktop 'patch and update' routine, but in slow motion.

Granted, things have gotten more complicated since the 1990s, but with all the advances in usability, no one has really thought about 'building something to last'. And the sad thing is you really can't. Because there's not a really big market for enterprise level networking equipment because of those service and license costs. So instead of third world nations buying up this equipment and progressing into civilization, we bury the silicon or recycle it for the gold.

And thus is the state of things from my opinionated view. :D
 
True, but Meraki wasn't going anywhere, and nor is their system as it's been purchased by Cisco.

It is still a major issue. when you have a product that relies entirely on cloud control, you are at the mercy of the company which knows that you are a captive consumer base. They can easily decide to charge for previously free functions, or remove functions, or simply stop supporting older products. e.g., look at the skydog router.

You also have many additional points of failure as if anything goes wrong within the path of your access point, all the way across the various routers and switches, up to the servers owned by the company, then your use of the product can be heavily crippled even for LAN based functions.

It has not happened in a large scale yet to really get the public to wake up, but one day when you see a more popular company, dropcam, nest, and other similar items, decide to either drop support for old devices, or kill the service in pursuit of other business ventures, or simply decide that everything will be behind a pay wall, and that pay wall will be 10 times more expensive than previous pay items.

Remember, the service does not have to completely shut down to effectively brick your device, they just have to price you out of using the product that you have already paid for, or decide to exercise some direct control over planned obsolescence.
 
Last edited:
It is still a major issue. when you have a produce that relies entirely on cloud control, you are at the mercy of the company which knows that you are a captive consumer base. They can easily decide to charge for previously free functions, or remove functions, or simply stop supporting older products. e.e., look at the skydog router.

You also have many additional points of failure as if anything goes wrong within the path of your access point, all the way across the various routers and switches, up to the servers owned by the company, then your use of the product can be heavily crippled even for LAN based functions.

It has not happened in a large scale yet to really get the public to wake up, but one day when you see a more popular company, dropcam, nest, and other similar items, decide to either drop support for old devices, or kill the service in pursuit of other business ventures, or simply decide that everything will be behind a pay wall, and that pay wall will be 10 times more expensive than previous pay items.

Remember, the service does not have to completely shut down to effectively brick your device, they just have to price you out of using the product that you have already paid for, or decide to exercise some direct control over planned obsolescence.
I actually never heard of skydog. Interesting read. Although I would have seen that coming a mile away. But yes, that's always a risk. How much of a risk is what you as the designer of a network have to assess.

Yes, they could be. But what will scare you to death is that most enterprise networks are already on those systems. Just look at the Sonicwall, Watchguard, and Fortigate routers out there--not only are they pricey, they won't work right without a software and support contract.

Yes, that can happen to any company anywhere, and it does. But then the consumer will do what they do in a capitalist economy--either pay up or find the next best value. Or they'll figure out how to hack it and use what is already there, like what happened to the HP Touchpad.

For businesses, it would come down to being quite upset at their previous decision, and then some emergency planning to see what can be put in place to hold the fort together. But this happens in businesses every day--from the fax machine that you can't buy thermal paper for anymore or the 20 year old laser printer that no longer has toner available. Change unfortunately comes to us all.
 
It seems that all user control is done via the cloud, which means that if they go out of business, or decide to kill the servers, or drop support, then your produce will become an overpriced plastic brick.
Where did you get this information? There is no mention of this in eero's FAQ.
 
I emailed the customer support.

"All device management (aside from initial set up) is done using our cloud services."

Basically just like skydog, you set it up just enough to get it connected to their servers before you can do anything else with the product.

The issue with many companies in this industry is they are not using the cloud to enhance a product, instead, they are using the cloud as a potential revenue stream as well as direct control over planned obsolescence.

For the cloud to enhance a product, it must not become a point of failure for the product. Cloud management for these products is simply unnecessary, especially as the only way to manage the device. Products like this can easily have fully local management over all functions, with additional cloud management for easy remote management. What is becoming more common is companies splitting the controls for their hardware where the back end code is housed on the consumer equipment, and the GUI front end stuff is housed on a remote server.

Unless consumers become more aware about how the devices they are purchasing are functioning, they will only end up burned in the future when they get blindsided by cloud planned obsolescence.

PS while firmware modding can help, some companies go out of their way to remove the ability of the user to ever modify the firmware, e.g., look at dropcam, forcing you to rely on the cloud is the pillar of their cloud revenue stream and they made damn sure that you would not easily get local access and turn it into an IP camera.
 
Last edited:
Indeed. Troops' mesh technique has a controller used to admin but not operate. Aruba is the opposite.
 
I emailed the customer support.

Unless consumers become more aware about how the devices they are purchasing are functioning, they will only end up burned in the future when they get blindsided by cloud planned obsolescence.

I agree 100% with this comment, that is why I don't use any of the current crop of smart devices since they don't function properly without their cloud counterpart. And the cloud counterpart is usually used to generate revenue for a gigantic corporation who's motto used to be "Don't Be Evil".

I was seriously looking at this mesh solution but it bring back the bad taste that Cisco did with their cloud connect configured Linksys routers and forced everyone into the cloud with an automatic firmware upgrade? There are some devices inherently need back end cloud services like video game consoles, streaming devices etc. But a router and other "appliances" that have stand along functionality doesn't need it.
 
Last edited:
Like it or not, we all depend on things beyond our control, "the cloud" being only one of them.

Every time we buy a product, we are taking a risk, with or without cloud baked in. Has the product been designed well? How long will it be supported? What happens after the vendor stops supporting it?

The key issues are foreknowledge and choice. Is the vendor sharing all the information I need to make an informed choice? Can I choose another service if one goes belly up or moves in a direction that I don't like? Or have I invested in a dead end and have to chalk the cost up to learning.

Never, as in "I'll never do X", is a long time, especially in the technology business. One thing I've learned watching the consumer networking business be birthed and grow over almost 20 years now is to keep an open mind.

Today's startup can become tomorrow's giant. Companies known for cheap knockoffs can move to the forefront of their industries. Anyone remember when LG was Goldstar and known for making cheap toaster ovens sold at Walmart?

Keep an open mind. Keep learning.
 
While you are taking a risk with any product you buy, it is important to also consider your control. for example, I bought a few cheap acer displays, after the warranty, all of them failed (bought 3). But for about $13, I repaired all of them by replacing the crappy capxon branded capacitors, with higher quality capacitors.

A product that does not rely on a company keeping a random server running, will be able to run almost indefinitely as long as you are willing to repair it when needed.

If you wanted today, you could hook up a colecovision from 1982, and it will still work today (not sure who would want to), but what if it relied on Coleco Industries to keep a server running in order for you to use the product?

This control circuit for an extremely old (must be around 50 years old) alarm system, still works (it fairly recently decided to remove the box, but a quick test caused the alarm to ring).
https://www.flickr.com/photos/razor512/14853209863/sizes/o/

Once you make the product completely dependent on a remote server, then the product will only ever last as long as that server, regardless of how well you maintain the client side hardware. Like in my previous example, for those who own a dropcam, what will become of your camera when the company goes out of business, or decides to no longer run those cloud services? What happens if you got it to record footage for security purposes, and the company increases their already insane $10 per month per camera, to something like $500 per month per camera?

Overall for cloud reliance, you are giving up 100% of all future reliability of the product, as it can be effectively bricked on the whim of a corporate executive.

Cloud simply does not allow for consistency or reliability. Your user experience is completely out of your control, as they can change, remove, charge for what was previously free, change prices for paid functions, or decide to no longer support older models.
 
Last edited:
I understand your points. All I'm saying is make a buy decision based on a product's risk / reward. Sometimes the risk is worth it.
 
The issue with many companies in this industry is they are not using the cloud to enhance a product, instead, they are using the cloud as a potential revenue stream as well as direct control over planned obsolescence.

For the cloud to enhance a product, it must not become a point of failure for the product.
Well said. You've nailed all the problems with the cloud offerings with this one statement.
If you wanted today, you could hook up a colecovision from 1982, and it will still work today (not sure who would want to)...
Ummm...these guys do:
http://atariage.com/forums/forum/115-colecovision-adam/
:D
 
What "cloud" dependencies do I NEED and have?
  • Internet Access
  • eMail


What cloud dependencies do I use but could do without (and used to do so)
  • News
  • On-line banking. (write a check)
  • Retail shopping (eBay, Amazon, et al)
  • Shipment tracking


I don't agree that vendor-self-serving cloud control just has to be for us. I wouldn't buy a networking product that cannot work and can only be administered only via the internet, or one that allows the vendor to admin remotely without an opt-out by me.

The worst of these is the electrical utility industry bribing legislators to allow the utilities to force time-of-day and rate-tier pricing upon us. And we rate payers have to even pay for the wireless meters at our homes to enable all of this. Outrageous.
 
Last edited:

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top