Why Hyperconverged For Small Business
-
@dashrender said in Why Hyperconverged For Small Business:
@pete-s said in Why Hyperconverged For Small Business:
@carnival-boy said in Why Hyperconverged For Small Business:
I'm not talking about HA. Just plain old non-HA environments.
However, with the ability to run some, or all, environments on a single host if another host fails. But you don't need to double the resources, as it is generally acceptable to run a slower environment for a few days.
That's manual HA with caveats.
Sure, it might be the best thing is some cases. Overconsolidating and putting all your eggs in one basket is not always the best.
But even if you get away with less than double the hardware you still need more than with just one host. So the hardware is going to be more expensive, the licensing of hosts and guest VMs is going to be more and energy is going to cost more.
I recall previous discussions around the eggs one basket thing. It doesn't really apply to most Small Businesses - why? because all of these services are generally needed. If one is down, the business is down, or at least crippled so much that those remaining don't matter. So putting everything on a single server isn't this huge risk that some think it is, because if the main app is dead, who cares about the rest.
Maybe, maybe not. The idea would be that you can run the main app on the other host and stop some of those less critical apps if needed to make space for the important stuff.
-
@pete-s said in Why Hyperconverged For Small Business:
@dashrender said in Why Hyperconverged For Small Business:
@pete-s said in Why Hyperconverged For Small Business:
@carnival-boy said in Why Hyperconverged For Small Business:
I'm not talking about HA. Just plain old non-HA environments.
However, with the ability to run some, or all, environments on a single host if another host fails. But you don't need to double the resources, as it is generally acceptable to run a slower environment for a few days.
That's manual HA with caveats.
Sure, it might be the best thing is some cases. Overconsolidating and putting all your eggs in one basket is not always the best.
But even if you get away with less than double the hardware you still need more than with just one host. So the hardware is going to be more expensive, the licensing of hosts and guest VMs is going to be more and energy is going to cost more.
I recall previous discussions around the eggs one basket thing. It doesn't really apply to most Small Businesses - why? because all of these services are generally needed. If one is down, the business is down, or at least crippled so much that those remaining don't matter. So putting everything on a single server isn't this huge risk that some think it is, because if the main app is dead, who cares about the rest.
Maybe, maybe not. The idea would be that you can run the main app on the other host and stop some of those less critical apps if needed to make space for the important stuff.
So you have two servers.
1- main app
2 - other less important stuff1 dies - now what? restore from backup on 2 and launch it? or have the main app always replicated to 1?
These are definitely options, but you need to look at the total costs of those to make sure it's business worthy.
If you're already going to the point of having two hosts - chances are that the extra RAM/storage isn't really that much to just make a sudo-HA setup (i.e. replicate VMs between hosts, manually power on in case of host failure)
-
@carnival-boy said in Why Hyperconverged For Small Business:
I've never really understood the desire for a single over double host.
All about money. If the value of the uptime is less than the cost of the second host, you don't get a second host. That's lost money.
These days, most people (most just being over 51% mind you) can shift the most critical workloads to cloud or desktop or some other holdover temporarily while waiting for a part.
I would say that 90% of my customers don't need a second server, and wouldn't see a single dollar of benefit to it if they had one, because they have other ways to keep the workloads running while the server gets repaired or replaced. And I've almost never had an issue getting parts under SLA because if they don't meet the SLA we get money back and they don't want that to happen. I've done literally thousands of HPE parts replacements (and other vendors too) and I can't remember (although I am sure that they have) them missing their SLA. Generally they would have parts in minutes rather than the four hour SLA period (NYC and Dallas are amazing for parts availability, of course.)
-
@carnival-boy said in Why Hyperconverged For Small Business:
Whatever contract I've had with HP has never prevented them saying "Sorry, your critical part is stuck at the port in Holland and there is nothing we can do".
DOesn't prevent them, of course. But that doesn't mean that you need to spend a fortune buying a second server to protect against it, either. All comes down to business needs. It's a risk / reward cost analyses for the given situation - which includes the company, the workload, the hardware, the vendor relationship, the age of parts, the ability to move to cloud, etc. etc.
But while it doesn't prevent that problem, I've also rarely had that problem lie in the critical path, either.
-
@carnival-boy said in Why Hyperconverged For Small Business:
And I'm not sure the costs of two hosts are significantly higher - you're still looking at the same amount of processing power, memory and storage, which are the main costs. Plus licensing, but that is variable depending on what you're running.
Well let's assume small server. Maybe $5000 in hardware. Licensing might be another $5000.
A second server is unlikely to double that, but is likely to double the hardware portion if you license nothing ahead of time. That's a big number to most companies. Might be a sensible number, but still large. Large enough to need consideration.
In most of the SMB, an additional $5K that exists only for rapid disaster recovery is a big figure. Especially when it is solely protection against hardware failure and doesn't address software, power, site, human and other emergencies which are the bigger factors. It's purely protection against prolonged hardware failure - something that typically I can overcome in 24 hours by calling xByte and expediting a server on the rare (i.e. has never happened to me) chance that we get stuck in that scenario.
WHile most small companies might be able to afford that, most small companies would struggle to justify the extra expense against its proposed value. But for some, it's also a no brainer. So it all depends.
-
@carnival-boy said in Why Hyperconverged For Small Business:
I'm not talking about HA. Just plain old non-HA environments.
However, with the ability to run some, or all, environments on a single host if another host fails. But you don't need to double the resources, as it is generally acceptable to run a slower environment for a few days.
That's true. Often you can get by with less. Maybe a lot less. Maybe only recover a few key workloads and not others (email yes, ERP no, for example.) But if you can do that, you can often skip having the server, too. Just use someone's desktop temporarily (we actually do this a LOT), or cloud, or whatever. And if you go far enough to actually buy a full server, often HA is so little more expense that it is worth going all the way to being able to load balance and fail over.
-
@dashrender said in Why Hyperconverged For Small Business:
@carnival-boy said in Why Hyperconverged For Small Business:
I'm not talking about HA. Just plain old non-HA environments.
However, with the ability to run some, or all, environments on a single host if another host fails. But you don't need to double the resources, as it is generally acceptable to run a slower environment for a few days.
Maybe yes, maybe no - at minimum you'd need double the storage, otherwise you can't the the workload from the down server onto the remaining one.
Also - is your plan to replicate the two hosts to each other so the data is available on the second host in case of failure? or is the plan to restore from tape?
I guess I just don't see that as really viable unless you go the full'ish HA route.
Double the storage only if recovering all workloads. He is saying that maybe some of them just don't matter during a recovery situation until the original host is repaired. So potentially it might be a smaller set of storage.
-
@dashrender said in Why Hyperconverged For Small Business:
While not impossible, I've only lost one server mobo in 20+ years, never lost a RAID controller or RAM. Scott has more experience here with 1000's of servers on wallstreet - but if memory servers (and JB will say it never does me) Scott experience there is still very low, likely to the point of not really worrying about it. But again, it depends on your situation.
We lost hardware willy nilly. A drive daily, memory sticks monthly, RAID controllers monthly, motherboards quarterly, power supplies monthly.... as percentages it wasn't bad. In raw numbers, it was a constant thing to deal with. But we always were able to get replacement parts fast and reliably. But, ideal locations, too. Not rural Iowa.
-
@dashrender said in Why Hyperconverged For Small Business:
One thing I haven't seen asked/talked about in this entire thread is - does the client still actually need their own servers? Can they put this in VPS? Like Vultr, etc?
Or at least during a recovery situation?
-
@dashrender said in Why Hyperconverged For Small Business:
If you're already going to the point of having two hosts - chances are that the extra RAM/storage isn't really that much to just make a sudo-HA setup (i.e. replicate VMs between hosts, manually power on in case of host failure)
So both points here are valid. Those points being...
- CB says that you need to consider a second host of partial capacity.
- Dash says that it is likely very little more to get to full HA / failover.
Both things are true. But there is a temptation to use #2 to ignore #1, which is false. It is definitely important to consider that if you are going to have a second host, does it then make it worthwhile to go all of the way to full HA? Yes, evaluate that. But don't assume it.
There are basically three options (not really, there are tons, but three big ones) which are single host, full HA double host, or limited second host. They have three different price points, three different value propositions, and all are valid options. You have to price them out and consider what protection each provides to determine which one (if any) is right in a given scenario.
CB's point is absolutely correct. You can't rule out having a second host only because full HA would be too costly. In some cases, a more limited second host, while far from free, might be a good decision while the additional cost of HA is not. But Dash is correct that this isn't a common scenario.
-
I step away for a day to work in the shop and come back to a group of mind readers. My next question was literally going to be how you handle situations where your best support is 4hr resolution but next business day for any resulting parts.
Assuming not a disk array failure, it would seem to depend on cost of one day of downtime and the cost to recover/regenerate any data since your last backup if your recovery method was via backups.
In the past suggestions have been to buy the same hardware and just swap the array over since, at least in cases of Dell, the array config is stored on the drives and the controller. This assumes matching controllers and firmware but still presents some risk.
You'd save the cost of the extra drives and complexity of HA but still have the capital waste of an idle asset. I think you limit excess licensing with open licensing. Is this even a semi-reasonable approach?
-
@woodbutcher said in Why Hyperconverged For Small Business:
Assuming not a disk array failure, it would seem to depend on cost of one day of downtime and the cost to recover/regenerate any data since your last backup if your recovery method was via backups.
If the issue is NOT an array failure, you can generally recover all of the data. There might be downtime, but not data loss.
-
@woodbutcher said in Why Hyperconverged For Small Business:
In the past suggestions have been to buy the same hardware and just swap the array over since, at least in cases of Dell, the array config is stored on the drives and the controller. This assumes matching controllers and firmware but still presents some risk.
This is universal and even software RAID (e.g. no controller) can do this (but with a tiny bit more effort.) Swapping controllers is a fundamental component of RAID. Since the controllers fail reasonably often, it is pretty critical.
-
@woodbutcher said in Why Hyperconverged For Small Business:
You'd save the cost of the extra drives and complexity of HA but still have the capital waste of an idle asset. I think you limit excess licensing with open licensing. Is this even a semi-reasonable approach?
Once in a great while, but almost never. There are a number of reasons... but basically in almost any case where it seems like idle assets are worth it (e.g. buying a second system but leaving it cold), you can make it warm for the same money and have way better results.
Remember a warm spare never requires additional licensing, only a hot spare. Warm spares are licensed the same as backup copies (because they are.) Because of that, the idea of the cold spare doesn't really hold much value.
-
However, all the concerns about licensing should raise red flags about your system. Yes, there are times that having something require licensing is okay or even desired. But you should always see it as a truly huge risk and one that you have to consider carefully. Licensing makes the most sense higher up the stack. For example, the application itself is the most likely object to justify licensing. The cabling is the least. Imagine how ridiculous it would be if your rack, desks, and cables required that you have a special license for each use!! It would be insane and you'd never stand for it (you might have to stand if your chair was licensed only to someone else's butt.)
Your hypervisor and operating system are closer to the cables, rack and desk than they are to the application in this way. These aren't components that you want to have at risk due to a need for licensing. Sometimes you have to, but it is rare and a "have to" would only be caused by an application and if an application puts you in that position for many companies that alone is a reason to question the viability of the developers behind said application.
At a minimum, something like ESXi introduces totally unwarranted complexity and risk and is something that can be quietly, transparently, removed and fixed during this process.
