Virt-manager: IDE disks
-
@scottalanmiller said in Virt-manager: IDE disks:
You are trying to install vProtect the agentless backup product for KVM?
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Trying to convert the ova to qcow2
Quicker than installing the app from scratch. -
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
-
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
-
@FATeknollogee said in Virt-manager: IDE disks:
Quicker than installing the app from scratch.
Is it? You still need to get to VirtIO in the end. And you want a reliable method for the future. I would not want to do it this way unless there was no other choice.
-
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
Yes, why are you looking at a crash-consistent "backup" tool?
-
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Quicker than installing the app from scratch.
Is it? You still need to get to VirtIO in the end. And you want a reliable method for the future. I would not want to do it this way unless there was no other choice.
For lab testing, this method "should" have been faster!
-
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
Yes, why are you looking at a crash-consistent "backup" tool?
Not too many other choices out there!
-
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Quicker than installing the app from scratch.
Is it? You still need to get to VirtIO in the end. And you want a reliable method for the future. I would not want to do it this way unless there was no other choice.
For lab testing, this method "should" have been faster!
I'm not sure that I agree. Pre-built appliances are nearly always a massive pain and they aren't a good process for testing end to end. So more work up front, more work later. I generally dislike them a lot. Faster they never seem to be.
-
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
Yes, why are you looking at a crash-consistent "backup" tool?
Not too many other choices out there!
What do you mean? There are endless awesome choices. This just isn't one of them. Why do you perceive a shortage of backup options?
-
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
Yes, why are you looking at a crash-consistent "backup" tool?
Not too many other choices out there!
What do you mean? There are endless awesome choices. This just isn't one of them. Why do you perceive a shortage of backup options?
List them out for me, please
-
Crash Consistent is the issue here, that I would see. When there are "real" backups that are full, rather than non-quiesced, why bother taking a backup if it isn't reliable? Reliability is the biggest factor in whether you consider a backup useful. It's not like you could use this and tell a client that you took a backup.
-
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Kinda, they have an appliance (prepackaged ova exported from ESXi), I was going to test it.
Crash consistent only, I'd not use that. Why are you looking at it?
Why am I looking at the application?
Yes, why are you looking at a crash-consistent "backup" tool?
Not too many other choices out there!
What do you mean? There are endless awesome choices. This just isn't one of them. Why do you perceive a shortage of backup options?
List them out for me, please
Veeam
Unitrends
StorageCraft
UrBackup
BackupPC
Amanda
Bacula
BackupExec
CloudBerry Backup
Cobian
Commvault
NetWorker
Spectrum Protect
Netbackup
YosemiteHundreds more, these are just the really big names.
-
In bold are the four that I would find most interesting and would only very rarely consider anything further. They range from free to pricy, self supported to enterprise support, and cover basically any possible scenario. Beyond those, I've had a lot of good luck with Netbackup in the enterprise.
-
@scottalanmiller said in Virt-manager: IDE disks:
In bold are the four that I would find most interesting and would only very rarely consider anything further. They range from free to pricy, self supported to enterprise support, and cover basically any possible scenario. Beyond those, I've had a lot of good luck with Netbackup in the enterprise.
Should have said, I prefer agentless!!
-
@FATeknollogee said in Virt-manager: IDE disks:
Should have said, I prefer agentless!!
Which, in this case is the same as saying "I prefer it not to work."
Agentless isn't an option here, period. Beyond that, why you have a "preference" for this kind of thing is an additional problem. IT should not have preferences, we should want proper solutions, however they work. Desiring a specific way of doing it, that can't be done, is an emotional mismatch.
-
@scottalanmiller said in Virt-manager: IDE disks:
@FATeknollogee said in Virt-manager: IDE disks:
Should have said, I prefer agentless!!
Which, in this case is the same as saying "I prefer it not to work."
Agentless isn't an option here, period. Beyond that, why you have a "preference" for this kind of thing is an additional problem. IT should not have preferences, we should want proper solutions, however they work. Desiring a specific way of doing it, that can't be done, is an emotional mismatch.
I hear you loud & clear, but, I still prefer agentless.
-
Since KVM doesn't support working agentless today (outside of very specialty systems like Scale HC3), you are at a starting point of having eliminated all working backups and only looking at a category that doesn't work. So naturally it will feel like your choices are limited, because there are literally zero.
But this should not make you feel that the choices are limited, the proper reaction is to step back and say "I'm doing something wrong, I'm not looking at the goal (to protect the servers), I'm stuck in the weeds of an emotional "want" rather than a business "need"."
-
@FATeknollogee said in Virt-manager: IDE disks:
I hear you loud & clear, but, I still prefer agentless.
Right. So you've identified the problem - an emotional breakage. You have to fix that, period. It's not a viable reaction. You need to step back and figure out why an emotion is driving you rather than reason and goal orientation.
Everyone has their emotions and preferences, but there is no place in IT for those emotions to creep into our decision making. None. The moment we've done that, we move from being IT pros into being purchases in a consumer process.
So in IT, one of our most important skills is learning to set the emotional preferences aside and focus rationally on goals and logical decision making.
-
I love the idea of agentless, it's great. But it's only an idea in this case. And one that honestly, isn't important at all. Sure, conceptually it is neat, but that's all that it is. From a business or technical perspective, it's just a pointless thing that only sounds cool because we improperly use the term "agentless" even though there is an agent. Even the title of the category is engineered to invoke an emotional reaction, and it works.
-
So, now that you know that there isn't an option for a working agentless backup for KVM... do you want to go with a non-working backup system that you can't claim to really be a backup, or do you want to pursue something that can actually take what is considered a backup by the industry to protect the workloads? What is the end goal that you are trying to achieve?
I realize that in a lab, playing with anything is fun. But this would be considered a waste of resources to focus on something that conceptually can't be deployed in any real world scenario.
