Why Hyperconverged For Small Business
-
This doesn't sound like they need a 2nd host to me. Is management not happy with a 4 hour part replacement warranty on a single server?
If they insist on at least 2 hosts, I think you're on the right track. No need for ESXi.
Edit: fixed bad grammar
-
That was another thought as well. I don't think their existing 2 node ESXi system is n+1 anyway so they essentially are already in this position, but maybe without extended warranty. Hard to tell as I am just digging in and their internal knowledge is not great.
The concern with that approach would be minimizing data loss. Primarily transactions in the ERP, though the volume of these transactions is low relative to other companies. But I would guess with proper backups at the DB level, this could be minimized as well if they had to recover via a backup.
With redundant PSUs, a healthy RAID setup, incoming power filtering, etc, the failure isn't likely to be a server death sentence. The issue (RAM, CPU, MB, etc.) would be resolved, fire back up the server and go back to making money.
-
@woodbutcher said in Why Hyperconverged For Small Business:
Is there truly a case for a hyperconverged infrastructure for a small business?
Not if you're looking for high availability.
Most small businesses lacks a lot of things to get a fault-tolerant high availability system. You need to look beyond the servers themselves because there are a lot of dependencies that also needs to be highly available. Like network, power, cooling etc.
Look at a commercial datacenter. It's usually redundant all the way through, from one end to the other.
-
@woodbutcher said in Why Hyperconverged For Small Business:
Is there truly a case for a hyperconverged infrastructure for a small business?
Yes, need for HC is based on your need for high availability which has little to do with company size. ALmost no one, of any size, actually needs HA. Large firms tend to have lots of politics and its easy to claim you need HA because there are big budgets and people like to feel important. But very rarely do they actually have a business need for it.
Small companies still rarely need it either, but have a harder time just throwing money at it so are less likely to do so.
-
@woodbutcher said in Why Hyperconverged For Small Business:
When I say small, I'm saying a company with 8 virtual machines and operating schedule of 12 hours per day, 5 days per week. While cost per hour has not been nailed down specifically, a couple hours of unplanned downtime is not something that would cause a lot of pain (assuming it is not a regular occurrence).
This definitely doesn't sound like a need for HA. The opposite, dramatically.
Very loose rule.... if you need HA you will know. It'll be so obvious that you'll never need to ask. If you have to ask, the answer is no.
-
@woodbutcher said in Why Hyperconverged For Small Business:
Currently they run ESXi on two slightly aging nodes and are looking to upgrade. The current proposed solution is a three node hyperconverged system with pedestrian specs and a $70k price tag. To me this doesn't pass the smell test.
Neither ESXi, nor HC meets the smell test. Having two nodes without HC already is a problem. Why two nodes if not HC? HC is the only normal reason at that size to have that many.
Why a THREE node HC, only two are needed on the rare chance that HC is needed at all.
-
@woodbutcher said in Why Hyperconverged For Small Business:
I feel ESXi is already an overkill solution here.
No, never think of ESXi as overkill. It's underkill. It's oversold. As technology it is less than you should have for free. If you paid one penny it is too much as better options are free already (including HC.)
ESXi is more of someone got sold a scam (in this case, ESXi is fine tech on its own), not that someone got sold "extra". There's no extra here, just not as much as there should have been.
-
@woodbutcher said in Why Hyperconverged For Small Business:
Pair this up with Starwinds vSAN to improve the reliability. With only two nodes you could avoid the 10G switching gear and just go direct connected.
This highlights some confusion. At your size, Starwind is the only viable HC option. When you said you got a price on HC, Starwind is really the only option until you need more nodes then Scale joins it. But at your size, Starwind is literally the only game in town other than doing the totally free options like Proxmox.
So basically you are saying that you can do two nodes and HC instead of.... doing HC, lol.
IF you are going with two nodes for reliability, then either stand alone or HC are the options. Given the likelihood that you have legacy apps (that ERP is likely) then HC is needed to make up for the crap apps. If you have good, modern apps, there is no value to HC whatsoever (or any HA at the platform)
-
Basically everything in this discussion is covered in my new book that came out today, by the way!! It's not about Linux but just general SA.
-
@woodbutcher said in Why Hyperconverged For Small Business:
These hyperconverged systems say you can do more with less hardware and cost but I am not seeing it. Even when trying to look at the fully loaded cost over the life of the system. What am I missing here?
HC does more with less. That is absolutely true and can be easily proven. WHat is not true is that paying for HC is going to get the same results. HA is free and low cost, if you want it to be.
SO just as Linux can be free and expensive, the benefits (and caveats) are universal. But the value depends almost entirely on how much someone convinced you to pay for it.
HC is the absolutely only sane way to do HA. If you need HA, it's untouchable 99% of the time. Far and away the best. BUT, if you look at garbage products (Nutanix, we're looking at you) then it doesn't matter than they are selling a good architecture, it's still a shit vendor with a shit product.
So three basic facts to keep in mind:
- Almost no one of any size needs HA.
- HC is the far and away best general case way to do HA.
- Bad products are always bad no matter how good one single under the hood aspect of that is.
Comparison: You might believe paper books are better than digital books. But a terrible story written on paper is still a terrible story. But that doesn't make digital books more pleasant to read.
-
@travisdh1 said in Why Hyperconverged For Small Business:
This doesn't sound like they don't need a 2nd host to me. Is management not happy with a 4 hour part replacement warranty on a single server?
If they insist on at least 2 hosts, I think you're on the right track. No need for ESXi.
Right, we should back up and start from the very beginning. Why is there ESXi (we know the answer... it's a mistake) and why are there two hosts (we think we know the answer... it's a mistake) and if we start over, dollars to donuts says that KVM or Xen (Hyper-V is kinda gone now, in any meaningful way) are the correct answers and probably on just a single host.
-
@woodbutcher said in Why Hyperconverged For Small Business:
That was another thought as well. I don't think their existing 2 node ESXi system is n+1 anyway so they essentially are already in this position, but maybe without extended warranty. Hard to tell as I am just digging in and their internal knowledge is not great.
Easy test...
If you need N+1, you should have HA. If you need HA, it has to be HC.
So it all comes down to answering "is N+1 needed"? Chances are, no. It isn't now. Nor are the most reliable setups being used. So nothing here suggests HA is even on the radar. Going down to a single host will lower cost and, we assume, increase reliability too!
-
@woodbutcher said in Why Hyperconverged For Small Business:
The concern with that approach would be minimizing data loss. Primarily transactions in the ERP, though the volume of these transactions is low relative to other companies. But I would guess with proper backups at the DB level, this could be minimized as well if they had to recover via a backup.
Unless you have DB level HA, nothing in your current set up (or in the HA setup!!!) will protect against transactional losses. Only proper database protection does that and nothing being discussed here touches on that.
-
@woodbutcher said in Why Hyperconverged For Small Business:
With redundant PSUs, a healthy RAID setup, incoming power filtering, etc, the failure isn't likely to be a server death sentence. The issue (RAM, CPU, MB, etc.) would be resolved, fire back up the server and go back to making money.
Even on Wall St. that is how they handle it because single servers have better transactional integrity and better overall reliability costs.
-
@pete-s said in Why Hyperconverged For Small Business:
@woodbutcher said in Why Hyperconverged For Small Business:
Is there truly a case for a hyperconverged infrastructure for a small business?
Not if you're looking for high availability.
Most small businesses lacks a lot of things to get a fault-tolerant high availability system. You need to look beyond the servers themselves because there are a lot of dependencies that also needs to be highly available. Like network, power, cooling etc.
Look at a commercial datacenter. It's usually redundant all the way through, from one end to the other.
Pete is spot on. The ability to do full HA by a company under 1,000 employees is nearly impossible and almost never valuable. The cost to actually do it is so absurd. But if you move to hosted, it can cost almost nothing. We have HA on things for pennies. The HA discussion is a big piece driving the ERP purchasing decision. WHoever chooses the ERP is the core decision maker about HA. Sounds like they likely decided HA was pretty far from a priority (which is reasonably accurate.)
-
@scottalanmiller said in Why Hyperconverged For Small Business:
@woodbutcher said in Why Hyperconverged For Small Business:
The concern with that approach would be minimizing data loss. Primarily transactions in the ERP, though the volume of these transactions is low relative to other companies. But I would guess with proper backups at the DB level, this could be minimized as well if they had to recover via a backup.
Unless you have DB level HA, nothing in your current set up (or in the HA setup!!!) will protect against transactional losses. Only proper database protection does that and nothing being discussed here touches on that.
I know you know Scott, but it bears repeating how HA usually works in pool/cluster of virtualized hosts.
The hosts in a HA cluster have storage in common but not much else.
When a host dies all data in each VM that has not been saved to disk (and replicated) is lost.
When the other hosts detects that one host is dead, all VMs that were running will now start and boot up again on other hosts. Since the storage is shared the VMs will have the same files as the ones that died.
The effect for the VM, and availability of the service the VM provided, will be about the same as killing the power to a server mid-operation and then power it up again.
That why it won't work reliably without transaction loss on a database.
-
@pete-s said in Why Hyperconverged For Small Business:
@scottalanmiller said in Why Hyperconverged For Small Business:
@woodbutcher said in Why Hyperconverged For Small Business:
The concern with that approach would be minimizing data loss. Primarily transactions in the ERP, though the volume of these transactions is low relative to other companies. But I would guess with proper backups at the DB level, this could be minimized as well if they had to recover via a backup.
Unless you have DB level HA, nothing in your current set up (or in the HA setup!!!) will protect against transactional losses. Only proper database protection does that and nothing being discussed here touches on that.
I know you know Scott, but it bears repeating how HA usually works in pool/cluster of virtualized hosts.
The hosts in a HA cluster have storage in common but not much else.
When a host dies all data in each VM that has not been saved to disk (and replicated) is lost.
When the other hosts detects that one host is dead, all VMs that were running will now start and boot up again on other hosts. Since the storage is shared the VMs will have the same files as the ones that died.
The effect for the VM, and availability of the service the VM provided, will be about the same as killing the power to a server mid-operation and then power it up again.
That why it won't work reliably without transaction loss on a database.
Right. In that sense we would call it a "crash consistent HA". Basically HA where the recovery is lightning fast, but only to a crash consistent level. Perfect for say a static website but bad for a database.
-
I agree with everyone else here - That's a setup that's overkill in hardware already (most likely) and potentially underkill in hypervisor (assuming ESXi Essentials).
Today you'd likely be fine with a single host running ProxMox, Xen or KVM and good backups.
The question is - how much down time can the company really handle?
When I first started here, I was told we could handle 7+ days of down time (we do everything on paper). At this point we're shooting for well under 1 business day. Of course since we using SaaS as our primary app is web based - if our ISP goes out - we are just hostage until that gets fixed... we're currently investigating cellular data backup. -
Why three? The "MSP" is trying to force these guys to purchase a solution from Scale Computing. As I understand it, they require three hosts. In any case, I feel this MSP is providing a solution that the MSP wants and is ignoring any and all business needs of their customer.
-
Thanks to all for the feedback. Some of these sales-only MSP organizations are making a mess of our industry. They are taking advantage of small companies and doing the exact opposite of what they say they are doing.
These same clowns are telling us you only need 16 cores worth of Server 2022 Datacenter to properly license these 3 proposed hosts. MS requires a min of 16 per host even if you only have 8 cores. Even if they have a license to re-allocate, they are still short.
They are just as bad at SQL licensing telling us you only need 2 cores to license a VM.
Unfortunately I was hired on after this process started and now have to try and unwind this mess before the purchase goes through. I worry I may be too late though.
I agree financially and technically, the single host approach is what makes sense here. I just have to put in the work to convince my new team that they have been fed a bunch of crap.
