Microsoft Launching Windows 11 on October 5

Brian_B said:
Well, the flip side of that is, you pay a cloud service a small amount, as part of a much larger overall client base, for several 9's in reliability and uptime.

If you wanted to achieve the same reliability/uptime on your own, it would likely cost you, unless you are a very large buisness as well, a significant amount more than it does to throw in with a cloud service. Data center grade isn't cheap or easy to maintain.

So if you are anything less than a Fortune 500 sized company, coming off the cloud would likely result in either higher expenses and/or less reliability overall.

You do, however, have the convenience of full transparency when you keep everything in house. It can lead to feeling more in control when problems crop up. You aren't just told "It's down, we expect it to be back in XXX hours" twiddling your thumbs while your company sits around idling a lot of man power and production based on lack of IT.
Click to expand...

Yea it's a balancing act. But the problem is no cloud vendor touches 5 9's reliability. And for my company that's what my team delivers so we are safe for now.
 
Grimlakin said:
Yea it's a balancing act. But the problem is no cloud vendor touches 5 9's reliability. And for my company that's what my team delivers so we are safe for now.
Click to expand...

So, I'm not professionally in an IT capacity, but I do have a rack and a few servers in the house and maintain small business-like type of services in a VM/Container server for the household.

I like to think of them as my "home production servers".

I'm not quite down with the lingo (5 9's is what, 99.999% uptime?)

I don't know if I'd quite achieve that, but the only times I've had unplanned downtime has been when I've had to move to a new house (and hopefully now that we own a place, that will be over for a while.)

I have had some planned downtime (usually scheduled when no one is around the house) to do hardware and software upgrades) but that's about it.

I'm sure once you do this **** in an actual business with larger numbers of users the complexity goes way up over my little system, but then you'd think you also have evenings and weekends to plan downtime for upgrades and fixes, which should make things easier.

Anyway, I'm just guessing, because I have no personal experience at all. I'm pretty sure my home setup rivals or exceeds many small to medium sized businesses, if the horror stories I have seen on various IT groups are anywhere near as common as they appear, but I am no corporate IT department :p
 
There can be external causes to outages like loss of utility power or loss of ISP, not to mention “forced update” that hits on an unplanned basis, so it’s not all just internal equipment related.
 
Zarathustra said:
So, I'm not professionally in an IT capacity, but I do have a rack and a few servers in the house and maintain small business-like type of services in a VM/Container server for the household.

I like to think of them as my "home production servers".

I'm not quite down with the lingo (5 9's is what, 99.999% uptime?)

I don't know if I'd quite achieve that, but the only times I've had unplanned downtime has been when I've had to move to a new house (and hopefully now that we own a place, that will be over for a while.)

I have had some planned downtime (usually scheduled when no one is around the house) to do hardware and software upgrades) but that's about it.

I'm sure once you do this **** in an actual business with larger numbers of users the complexity goes way up over my little system, but then you'd think you also have evenings and weekends to plan downtime for upgrades and fixes, which should make things easier.

Anyway, I'm just guessing, because I have no personal experience at all. I'm pretty sure my home setup rivals or exceeds many small to medium sized businesses, if the horror stories I have seen on various IT groups are anywhere near as common as they appear, but I am no corporate IT department :p
Click to expand...

I work for a company that delivers a service in the life safety Category, we do NOT do planned outages. We build with a level of redundancy I would call 16x. For us to suffer an actual services complete outage it would take a 16x type of failure.

So as an example, each of our physical servers has connections to two network switches over independent network cards that feed to 2 core switches with 2 or more hard line internet connections from different vendors. Beyond that we have our servers in groups so if a 'group' of servers go's down a second at that site takes over. Then we have multiple sites with the same sort of build methodology each since 'group' of servers able to handle this specific load for the entire company.

So in essence with each server in a group having 2 feeds for power and network and computation and then having that replicated with another server at the same site gives us a ton of fault tolerance.

If you add in multiples for dedicated ESXi hosts on top of that so we can shuffle non hardware servers between hosts seamlessly you add another multiple so 32x? Lets just say we take the lives of our customers very seriously and spend and design to make as sure as we reasonably and really beyond reasonably can that they are served no matter what.

That isn't to say your service you provide your family isn't of comparison in many ways it is. It's just the economics of scale and load that come into play.
 
Grimlakin said:
I work for a company that delivers a service in the life safety Category, we do NOT do planned outages. We build with a level of redundancy I would call 16x. For us to suffer an actual services complete outage it would take a 16x type of failure.

So as an example, each of our physical servers has connections to two network switches over independent network cards that feed to 2 core switches with 2 or more hard line internet connections from different vendors. Beyond that we have our servers in groups so if a 'group' of servers go's down a second at that site takes over. Then we have multiple sites with the same sort of build methodology each since 'group' of servers able to handle this specific load for the entire company.

So in essence with each server in a group having 2 feeds for power and network and computation and then having that replicated with another server at the same site gives us a ton of fault tolerance.

If you add in multiples for dedicated ESXi hosts on top of that so we can shuffle non hardware servers between hosts seamlessly you add another multiple so 32x? Lets just say we take the lives of our customers very seriously and spend and design to make as sure as we reasonably and really beyond reasonably can that they are served no matter what.

That isn't to say your service you provide your family isn't of comparison in many ways it is. It's just the economics of scale and load that come into play.
Click to expand...

Wow, Yeah, I've read about so called "High Availability" clusters in the past. Yours seems to take that to a whole new level.

And yes, very definitely, no ones life is at risk if any of my systems go down.

I have been on teams developing hardware and software for life sustaining Class 3 medical devices, and that was... well.. interesting to say the least.

One thing I've never quite been able to wrap my head around is how - in these highly redundant systems - the data is all kept in sync. I mean, it sounds like as if there is a backup system with a backup database if the main one goes down, but how is it maintained in a state of readiness such that it has all of the latest data, if it becomes needed?

Also, it would be interesting to do some sort of fault tolerant system like this that used MULTIPLE cloud services providers as backups for eachother.
 
Last edited:
Zarathustra said:
Wow, Yeah, I've read about so called "High Availability" clusters in the past. Yours seems to take that to a whole new level.

And yes, very definitely, no ones life is at risk if any of my systems go down.

I have been on teams developing hardware and software for life sustaining Class 3 medical devices, and that was... well.. interesting to say the least.

One thing I've never quite been able to wrap my head around is how - in these highly redundant systems - the data is all kept in sync. I mean, it sounds like as if there is a backup system with a backup database if the main one goes down, but how is it maintained in a state of readiness such that it has all of the latest data, if it becomes needed?

Also, it would be interesting to do some sort of fault tolerant system like this that used MULTIPLE cloud services providers as backups for eachother.
Click to expand...

Some use windows clustering to keep redundancy in sync with a common fileshare for changes to be propagated through. Others use custom solutions. I will say that ours is custom but we also backup the database of each server as if it is the only one religiously.
 
I haven't seen a start menu add. But I turn that crap off regardless.
 
I have used pretty much every communication/collaboration software out there, from webex, cisco, google meet, zoom, messenger, live, and many others I can't remember their names. But Teams take the cake, its probably not for everyone and maybe too much for casual users, but what you can do in teams you really can't do anywhere else.

Unfortunately NONE of my current devices are windows 11 compatible, so I'm happily stuck with windows 10 for the foreseeable future.
 
You must log in or register to reply here.
Become a Patron!
Back
Top