Burned by Eschewing Best Practices
-
@BRRABill My issue isn't the single DC.
My issue is it's running physically, it cost nothing to virtualize.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
I don't feel bad in the least..
"We had a user infected with the locky virus and now its spread to the shares on the fileserver.
Anyone encountered this before on a server side and what did you do to fix? We have backups the only problem this a physical server which is our only DC, WSUS, Exchanged DB, Sharepoint etc."
Single server running physically that is hosting DC services, WSUS, Exchange, Sharepoint and more....
I do not feel bad in the least.
Not even clicking through, but I would assume it is an old SBS server. The SMB was not doing virtualization in 2008 as a general rule. Yes some few businesses were, but not most SMB.
They should have just said SBS and not listed the components.
-
"only problem this a physical server which is our only DC, WSUS, Exchanged DB, Sharepoint etc."
Why is that a problem? Why can't they just restore the encyrpted folders, and leave everything else as is?
-
@DustinB3403 said:
My issue is it's running physically, it cost nothing to virtualize.
It costs them time to swing it over from physical to virtual. Time is always a cost.
-
@Breffni-Potter It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
The cost in time is so trivial that it shouldn't even be a question.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
@Breffni-Potter It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
The cost in time is so trivial that it shouldn't even be a question.
It depends on if they have IT staff, and how much data we are talking.
-
@BRRABill said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
@Breffni-Potter It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
The cost in time is so trivial that it shouldn't even be a question.
It depends on if they have IT staff, and how much data we are talking.
No it doesn't. If you don't have IT available, then hire someone. The amount of data simply adds time to convert. But that time to convert is often times far less to revert / restore a VM than it is to restore physical systems.
-
With a virtualized environment you can do both block level and file level restores much more simply. Hypervisor tools like Xen Orchestra and file level tools like Shadow Protect.
Gives you many rapid restore options.
-
@DustinB3403 said
The cost in time is so trivial that it shouldn't even be a question.
Actually it's not trivial if you don't know or understand it and there is a magic box in the corner of the office keeping everything going.
@DustinB3403 said:
But that time to convert is often times far less to revert / restore a VM than it is to restore physical systems.
But you are assuming the benefits are understood in advance of doing it? So if you don't understand the benefits, why would you go near virtual systems especially since the market name "VMWare" wants to sell you stuff, so you think it is higher cost than physical. Is this true? Not at all but it's easy for mis-information to creep in.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
Gives you many rapid restore options.
You get that with Veeam Endpoint also.
P2V can take days if you have a slow machine/network.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
With a virtualized environment you can do both block level and file level restores much more simply. Hypervisor tools like Xen Orchestra and file level tools like Shadow Protect.
You can do this with LVM. I can take a snapshot and export it to a file. Restore from that file later, or mount it and copy files off. Or just keep the snapshot and revert if I need to. Not knowing anything about Windows, you might also be able to do this with their dynamic disks or whatever it's called.
-
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
Gives you many rapid restore options.
You get that with Veeam Endpoint also.
P2V can take days if you have a slow machine/network.
Restoring Backup to Physical can also take days.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
Gives you many rapid restore options.
You get that with Veeam Endpoint also.
P2V can take days if you have a slow machine/network.
Restoring Backup to Physical can also take days.
So then this sentence has no weight.
It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
-
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
Gives you many rapid restore options.
You get that with Veeam Endpoint also.
P2V can take days if you have a slow machine/network.
Restoring Backup to Physical can also take days.
So then this sentence has no weight.
It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
Sure it does, I can revert a VM in a matter of minutes with Snapshots, or on the fly Disaster recovery. Where you literally turn on a VM that is 5 minutes old.
-
@DustinB3403 said in Burned by Eschewing Best Practices:
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
@johnhooks said in Burned by Eschewing Best Practices:
@DustinB3403 said in Burned by Eschewing Best Practices:
Gives you many rapid restore options.
You get that with Veeam Endpoint also.
P2V can take days if you have a slow machine/network.
Restoring Backup to Physical can also take days.
So then this sentence has no weight.
It cost substantially less time to virtualize "yesterday" then it does to restore physically today.
Sure it does, I can revert a VM in a matter of minutes with Snapshots, or on the fly Disaster recovery. Where you literally turn on a VM that is 5 minutes old.
You can do all of that with physical also. However, in that scenario you are already virtualized. So converting from physical to virtual "yesterday" takes just as long as restoring physically today because it's the same data. You have to get the data in the VM first, so that would take the same amount of time.
-
How are you making the assumption that restoring to physical is as quick as restoring a VM.
Let's use an example, physical server catches on fire, you need to completely replace it but have the ability to restore to dissimilar hardware.
So you have to, purchase hardware (wait for it to arrive, connect it up) and then restore it.
If you have a hypervisor that dies, you can send the VM's to a datacenter in a matter of minutes (if using a local provider) and have everything back up and running.
Or you could even have the dead Hypervisor in a HA mode, where the VM is live on the other host as soon as an issue is detected?
-
@DustinB3403 said in Burned by Eschewing Best Practices:
How are you making the assumption that restoring to physical is as quick as restoring a VM.
Let's use an example, physical server catches on fire, you need to completely replace it but have the ability to restore to dissimilar hardware.
So you have to, purchase hardware (wait for it to arrive, connect it up) and then restore it.
If you have a hypervisor that dies, you can send the VM's to a datacenter in a matter of minutes (if using a local provider) and have everything back up and running.
Or you could even have the dead Hypervisor in a HA mode, where the VM is live on the other host as soon as an issue is detected?
We aren't taking about restoring a VM. We are talking about converting from physical to virtual vs restoring to physical.
The person in that post doesn't have virtual, just physical. So the time to convert to virtual has to be similar to the time taken to restore from physical since it's the same data.
-
I think you missed the point I was making originally, which was if he was virtualized this wouldn't be as big of an issue that it is.
Which "oh shit we got crypto-locked at 12:42AM" OK revert the VM.
The conversion process is whatever time it takes, I even said that. But the time to recover a physical only system versus a Virtual one is way more time intensive.
-
But the time to recover a physical only system versus a Virtual one is way more time intensive.
But like I said I can do snapshots and file level restores physically also. Same amount of time as virtual.
-
So yes while being virtual would be best, saying "I don't feel bad at all" like they're some kind of moron is not something that should be said. As @JaredBusch pointed out, it's probably SBS. Hyper-V sucked back then. So you're left with paying for VMware or using a Linux system that you might not know.
