Hi,

For what reason did you remove the advanced schedule options, in the new backup format. If I want to make a image backup, I can choose for daily backup off course, and only "execute full backup" monthly, weekly or daily.

A synthetic full backup, is a full backup, that is re-using the storage that's already in the cloud. It's saving upload time, but in the cloud himself they put all blocks together as a new full backup. So 10 GB backup, next month again 20 GB, and next month same story 30GB.

Wasabi is having a minimum of 90 days storage, I had it set to a full backup every 45 day's and keep 45 days. So the storage needed to be 2 full, and all diffs. But I maybe wrong, now there one more full backup space needed in for 3th month, because you decided that we can't set other option then daily, weekly or monthly.

Hi David,

Thanks fo the reply, I found it. Perfect!!! I hope it will help others also, thanks an good job, koop on the going!

Kind regards,

"Use Backup Plan wizard to specify which files, folders or disks you'd like to exclude from the Hyper-V backup." — Denny

Will this include, excluding folders inside the virtual machines hard drives?

Fixed it already a other way, I created a Windows USB 10 installation disk and replaced the boot.wim file.

Cloudberry image based back-ups can be used on Hyper-V clients, but you will get some small problems. One of them is that I wasn't always able to restore the server, the Cloudberry recovery disk was very slow in lab environment.

But the most important part is that if you have a image based back-up you're HAL will have a other ID, then the new server where you will restore it. One of the most simple examples is that you get a new network card inside you're server and it's called #2. So also you're ID's will be changed, what can be very annoying when you try to restore a domain controller.

Beside the performance issues, it's always better to back-up from the host. It will send a snapshot message to the client, what will make a volume shadow copy and then create the back-up.

I'm also waiting a long time for Cloudberry to support at least synthetic full, but even better CBT and they are already supporting that for VMWare. CBT is the most efficient way and for MSP's it will be perfect if it is possible in Microsoft Azure, because as a CSP partner you can make just a little more money and let you're client host there back-up in their own Azure space.

Hi David, thanks and I understand all of them. But maybe my question was wrong, what I want to know.

1. If I have a full back-up
2. and lets say 10 block levels
3. and I delete the 10 block levels
4. Do a consistency check
5. Run a new block level back-up
- Will this new block level backup have all changes from the first full or do it still needs the 10 missing blocks?

This is just for a test case scenario.

No problem, good to know that it is in the planning.

Because I don't have the space at the moment for a new full backup... On my lab environment

CBT for Hyper-V and when possible other types of backups, it will save you to do a new full back-up.

Hi James,

Sorry I indeed mean onsite 1 time a week a full, but with 2 weeks retention. And monthly for the cloud storage, cloud storage is for the extra backup when onsite will faill. So if it takes more time, it takes more time. It's like a disaster, disaster recovery ;-)

I also tested a image back-up inside the VM, but it really takes a lot more to time to backup then backup on Hyper-V level. And at the moment there is no Syntatic Full Backup option for Hyper-V and there is no Synthetic Full for Azure.

I like Azure more, because I provide almost only Microsoft products to my clients, and as a CSP I also get a percentage back from Microsoft and I can create unlimited containers (buckets) in one account. The last nice thing on Microsoft is that when you buy Action Pack you get $ 100 Azure monthly credits, thats like around 6 TB of back-up.

Ah thats sounds promosing, so I need to setup chaining for 2,3 an 4?

The backup will skip the chain if it's not finished by the time the chained run is triggered.

As a rule, any plan will skip scheduled launch if it's still running by that time. — Matt

Maybe it's my English :-) But I understand here that it will not run if the backup is still running and the next scheduled needs to be started.

So for me that looks like I need to schedule the plans to run when i'm sure they are finished.

Is there a way to program the backups in a chain, so the next backup will start after a backup is finished? What I want to reach is to program backup tasks like:
1. Run: Backup system to network share (2 weeks retention. every 2 weeks Sunday full, every day block)
2. Run: Backup data to network share (2 weeks retention. every 2 weeks Sunday full, every day block)
3. Run: Backup system to cloud (4 weeks retention. every first Sunday of the month full, every day block)
4. Run: Backup data to cloud (4 weeks retention. every first Sunday of the month full, every day block)

- I don't want to use hybrid because I don't want to be depended to be the onsite (hybrid) is working
- I want to use different retention for onsite and offsite

Sorry I didn't ask the question right. Can I remove a single VM from backup and the storage when multiple VM's are backing up in this backup plan, or are they part of the total backup plans block levels?

I'm happy that I read this, because yesterday night I was almost going to install a Minio server on Windows in the night. So good to know that it is not compatible at the moment, but I hope it will be in the feature.

I will like to know if it's possible, also I can also setup a test server for it and share logs and help with it. Minio looks good, I also did some reading about Minio multi tenant. Will nice if Minio will be supported for Windows, or do you know some other S3 compatible software. I did some research but didn't find that much.

A suggestion for you, if you need to exclude drives now. The workaround i'm using now, is a lot far from ideal but it works. Some clients from me don't want to upload some shares to the cloud to save in the costs.

So what I do now is:
1. Having a VM with 3 VHDX files
- C: System
- D: Data
- E: Data no cloud
2. The virtual machine is created and mounting
- C: System
- D: Data
3. This virtual machine is backed up to the cloud
4. Drive E is a VHDX on a share \\%hyper-v-host%\vhdx$\%virtual_machine_name%\%virtual_machine_name%-E.vhdx
5. I have a script inside the VM to:
- Get Hyper-V host variable
-- Powershell: $VMHost = (Get-ItemProperty -path 'HKLM:\SOFTWARE\Microsoft\Virtual Machine\Guest\Parameters').Hostname
- Get Hyper-V client variable
-- Powershell: $VMClient = (Get-ItemProperty -path 'HKLM:\SOFTWARE\Microsoft\Virtual Machine\Guest\Parameters').VirtualMachineName
-- If you like more to open share host IP adres: $HVIP = ([system.net.dns]::gethostaddresses($VMHost) | Select-Object IPaddressToString).IPaddresstostring
6. Inside the virtual machine I create a scheduled task to auto mount the file: \\%hyper-v-host%\vhdx$\%virtual_machine_name%\%virtual_machine_name%-E.vhdx
7. Now only drive this is in the Hyper-V backup
- C: System
- D: Data
8. I backup drive E inside the virtual machine to a network share and because this drive is mounted inside the virtual machine, it;s not backed up by Virtual Machine backup

I know it's long workaround, but sometimes clients don't want to pay for a lot of cloud back-up. And when using image back-up inside the virtual machine is a lot slower (and I mean a lot) then backing up the virtual machines on Hyper-V host level.

Thanks for your request - we raise a priority of the feature with every new customer. Talking about your second wish, it is coming very soon - within CloudBerry Backup 5.9 that is expected in 1-2 weeks. — Julia

Version 5.9 is now available and I can't find such a feature you have promised above. — ds2

This workaround doesn't mean that we don't need the promised feature anymore, because when using Hyper-V replication the E drive is not there ;-)

I also see and tested "Backup Virtual Machine Edition (5.9.0.313)" and can't see any of this options.

Even when copying such amount of files locally such speeds are expected. Encryption and NTFS permissions might not slow the operation down on a small scale, but when dealing with large quantities of files these options affect the upload. — Matt

But what do you advise? Enryption is now set to AES 256, NTFS permissions are backed-up and compression is enabled. NTFS and encryption is needed and maybe no compression, but that will make the data be a lot more.

Yes, I already read that above. I only wanted to explain a scenario, for other users to understand the reason why the ability to restore NTFS permissions is important

But that sounds very good and will help to make this product more powerful, good job!

Image based back-up will speed this up, but then it will be perfect to have the possibility to restore file level from the image (whats already possible), with an option to restore NTFS permissions from the image based back-up. So we can only restore some effected data.

Let's say this scenario:
1. A server with about 1 or 2 TB of data, a file level back-up will take forever. So we make a image based back-up from drive D
2. Workstations are connected to the domain controller
3. A user logs in with roaming profile and folder redirections
4. The user is affected with e cryptowhere so files that have been affected:
- \\data.client.domain.nl\profiles$\%username%
- \\data.client.domain.nl\users$\%username%
- \\data.client.domain.nl\dfs$\groupdata shares where the user has rights to let say:
-- But the user don't have rights to:
--- \\data.client.domain.nl\dfs$\administration
--- \\data.client.domain.nl\dfs$\direction
5. So the only needed restores are:
- \\data.client.domain.nl\profiles$\%username%
- \\data.client.domain.nl\users$\%username%
- \\data.client.domain.nl\dfs$\groupdata
6. When having a file based back-up it's easy to restore the affected share
7. With image based back-up the complete drive needs to be restored, but users that have permissions to also have new files for that day:
- \\data.client.domain.nl\dfs$\administration
- \\data.client.domain.nl\dfs$\direction
8. So now we have a problem, there a affected files from point 5 and changed files from point 7

Doing a full disk restore and replacement will lose files from point 7, so we only want to restore files from point 5 "WITH" NTFS permissions.

So having the possibility to restore NTFS permissions from image based back-up will be awesome and make the product perfect for simulair scenario's.