Without any doubt, I find the hardened Linux repository Veeam introduces in Veeam Backup & Replication v11 one of the most fascinating new features to get my hands on. In the ever-escalating battle with ransomware and wipers, this is a very valuable option to have in your defensive arsenal. So, I grabbed the Beta 2 and got to work in the lab over the holidays to investigate and find out some details about the Immutability of Linux files on the Veeam hardened Linux repository.
Immutability of Linux files on the Veeam hardened Linux repository
I’s quite easy to find the file attribute “i” that marks a file as immutable.
attr -a -l attr -a -l
attr -a -l
Where is the information about the immutability actually stored? I mean, that “i” attribute is one thing but how do the Linux host and Veeam know from what time period this immutability is valid. In the end, the service has to clear it and know when to do this. Or is this only stored in the Veeam database or both?
How does it now from when till when a file must be immutable?
Digging around in the files and folders of the Veeam repository, I soon found the lock file “.veeam.x.lock” (see the green arrow in the image above) that is created by the veeamimmureposvc service. The owner is root, hence it is not created by the Veeam transport service. The veeamimmureposvc service is a local account with root access for managing the immutability. It only works locally and does not listen on any network port, hence it cannot be accessed remotely.
Let’s look inside to see if we can read something there?
When you open that file you will find it to be an XML file. Inside you’ll see the date and time stamp for every file in the backup chains for that job. That’s cool.
But there is more. When we run to look for extend file attributes we find that every Veeam created file has a one called user.immutable.until.
With that name, it is clear it can be of interest to us. If you look at what is in there, you’ll see it contains the date and time stamp for that file’s immutability period.
getfattr * -n user.immutable.unil
That I find interesting. Veeam saves the information twice. Is that for redundancy or as some sort of checksum? Maybe it also has to do with the fact Veeam backup files are transportable and self-contained so that information is stored as an extended file attribute.
So there you have it. A small piece of information on where the immutability information is stored. The most surprising thing to me was that it is actual stored twice.
I hope you fund this interesting. Poking around to figure out the how and what of things always helps me tremendously to learn and understand the technologies I want to work with. That leads to better decisions in design and implementation. It leads to both trust and confidence, which helps me decide where and when to leverage it. Finally it also, almost without, it is invaluable when supporting the technology.
Since diving into the Veeam Backup & Replication v11 Linux hardened repository I have started to use XFS in bite-size deployments to gain experience with it. One of the things that will certainly come in handy is extending a Veeam Repository XFS File System. In this blog post, I show to do that.
Mind you that I am doing this with a virtual machine on Hyper-V (Windows Server 2019) in the lab. Not every permutation of hardware and storage controllers you can find. But still, the procedure here will not differ that much.
Determine the size of the current disk.
Now take note of the bytes and sectors
sudo fdisk -l
Expand the disk
In the virtual machine settings I extend the virtual disk I want to grow with the required capacity.
In real life that might be you growing a raid controllers’ virtual disk by adding physical disks to the raid controller, you expanding the volume on the storage array or simply adding disks to the local server and adding them to the software-based raid solution you use.
The virtual machine will pick up the extra capacity right away. For our UBUNTU 20.04.1 OS to see it up we’ll need to rescan the SCSI busses for change. In a virtual machine, this can be done via rescan-scsi-bus.sh, available scsitools that will need to be installed if not there.
Use the -s options as that will really show the resized disks.
Note: What I did find is that if you just expand the disk and than extend the xfs file system, it also works. It seems to just work without rescanning the disk after extending it! The disks size in df -h will show this space then as well.
That was it. Short and sweet. There is not much to it once you know how to do it. One thing to remember is that you cannot shrink an XFS file system. So, as always, start smaller and grow when needed. Always leave spare capacity to work with when needed. Yes, even in 2021 this is advice to live by in the storage world. For Veeam this means that multiple smaller repositories or extents give you more wiggle room than fewer very large ones. Leave capacity in reserve, either in a spare repository/extend or unallocated. This, especially combined with a scale-out backup repository in Veeam will allow you to work your self out of most capacity pickles you might find your self in.
As I have started to use XFS in bite-size deployments to gain experience with it I wanted to write up some of the toolings I found to manage XFS file systems. Here’s how to check/repair/defragment an XFS volume.
My main use case for XFS volumes is on hardened Linux repositories with immutability to use with Veeam Backup & Replication v11 and higher. It’s handy to be able to find out if XFS needs repairing and if they do, repair them. Another consideration is fragmentation. You can also check that and defrag the volume.
Check XFS Volume and repair it
xfs_repair is the tool you need. You can both check if a volume needs repair and actually repair it with the same tool. Note that the use of xfs_check has been depreciated or is not even available (anymore).
To work with xfs_repair you have to unmount the filesystem, so there will be downtime. Plan for a maintenance window.
To check the file system use the -n switch
sudo xfs_repair -n /dev/sdc
There is nothing much to do but we’ll now let’s run the repair.
sudo xfs_repair /dev/sdc
The output is similar as for the check we did for anything to repair is basically a dry run of what will be done. In this case, nothing.
Now, don’t forget to mount the file system again!
sudo mount /dev/sdc /mnt/veeamsfxrepo01-02
Check a volume for fragmentation and defrag it
Want to check the fragmentation of an XFS volume? You can but again, with xfs_db. The file system has to be unmounted for that or you will get the error xfs_db: can’t determine device size. To check for fragmentation run the following command against the storage device /file system.
sudo xfs_db -c frag -r /dev/sdc
Cool, now we know that we can defrag it online. For that we use xfs_fsr.
xfs_fsr /devsdc /mnt/veeamxfsrepo01-02
xfs_scrub – the experimental tool
xfs_scrub is a more recent addition but the program is still experimental. The good news is it will check and repair a mounted XFS filesystem. At least it sounds promising, right? It does, but it doesn’t work (Ubuntu 20.04.1 LTS).
That’s it. I hope this helps you when you decide to take XFS for a spin for your storage needs knowing a bit more about the tooling. As said, for me, the main use case is hardened Linux repositories with immutability to use with Veeam Backup & Replication v11. In a Hyper-V environment of course.
A File-Level Restore in a hardened environment can trip you up. In Veeam Backup & Replication the ability to do file-level restores from virtual machine backups is a handy option to have. In secured environments with an isolated and protected Veeam backup fabric, this is not always a straightforward or fast process. Let’s look at this and make some suggestions for a solution or workarounds.
Veeam File-Level Restore in a hardened environment
Nowadays, it is a common best practice that the Veeam Backup & Replication environment is isolated and independent from the fabrics it protects. It is even more than a best practice and ransomware educated many of us fast and hard on the need for this.
In such an environment every server in the Veeam backup fabric solution is isolated, with their own credentials and not joined or in any way dependent on the environment it is protecting. Dedicated accounts are used to run the actual backup jobs and these are not used for normal administrative or operational tasks.
The majority of Veeam functionality works just fine in such an environment. But there are some things that need to be addressed. One such thing is a file-level restore (FLR). In a secured and isolated environment, an FLR job cannot leverage admin shares to gain access to the virtual machine to where files are being restored or access the data in the backup. To work around this Veeam can leverage PowerShell Direct by which it can securely restore the files into the virtual machine anyway.
But leveraging PowerShell Direct comes at the cost of speed. 7 MB/s is okay for a small number of smaller files. It becomes frustratingly slow when it comes to restoring many and larger files. Minutes turn into hours. In some cases, this just won’t do. In our case, the DBA who needed to do the database restore was not happy and the developers waiting for the restored database were not overly pleased either.
So what other options do we have?
Set up a secured file share in the protected environment
Below we will discuss the other options you have for a File-Level Restore in a hardened environment via FLR in Veeam Backup & Recovery. But first, we prepare a landing zone in the secured environment where we can copy restored files to. For this, we set up a restore share in the protected environment where a few admins that are responsible for backup and restores can write to. No one else can enumerate, read, or write to that share. When a file gets restored there they can copy it over to where the people that need it can access it. This helps protect and limit access to restored files to just the people that need to get to it. The accounts used to access that share from the isolated backup environment do not get admin permissions on the system where that share resides or anywhere else.
Providing a restore share in the protected environment is safe. You only allow certain admins to have access to it, no one else. Since the less trusted environment (the one we protect) allows access from a more secure environment (the backup environment) this does not expose the backup environment to extra danger. At no point does this allow any access to the backup environment from the protected environment.
Use the FLR wizard to copy the files
You can use the FLR wizard to copy the file locally in the secured Veeam backup environment.
From there you can copy it to the restore share in the protected environment. This is very fast as nothing is. On our network we get 230 MB/s.
But what if you don’t have the space for that locally? Not everyone can have a large local restore disk onWould it not be handy to copy it to restore share in the protected environment directly? Well yes, you can try in the wizard and it will ask for the credentials to access the file share.
But it will fail to read and restore the backups but this will fail with an “Access is denied” error.
You could work around it if you use the account VBR uses to actually make the backups. But that won’t do. For one those credentials should be guarded and not use by anything or anyone but VBR. Secondly, it doesn’t work anyway.
So are you out of luck? The good news is no, you are not, so read on!
Copy the files directly from the VeeamFLR mountpoint
If you don’t have the space to restore the file locally in the secure backup environment, there is another trick. When you look at the restore log you will see what the mount server is. That is the node where you can find the C:\VeeamFLR folder, which is where the content of the virtual Machine VHDX volumes is being mounted. This means that on that node you can navigate to the folder or files you want to restore.
Select the ones you want to recover and just copy them. Navigate to restore share in the protected environment. You just need to provide the credentials of the dedicated account with write access to that share. Those are the only rights you need on that share. The copy is again fast at 230 MB/s. That is a lot better than 7MB/s.
There is an issue with this, however. In this scenario, you need access to the server where the volumes of the virtual machine get mounted. Normally that would be the local repository where the backup files reside and guess what, you lock these down as much as possible and don’t log on to them unless needed. You could specify another, but then it needs to be a server where you have access to and it copies the data over the network, which is slower. With 10Gbps or more that’s not an issue but with 1Gbps, it can take a while with a lot of large files. Can we fix this? Yes. with Mount to Console!
Mount to Console to the rescue
But hey, this would not be Veeam if they did not have yet another option on their sleeve! You can use the Mount to Console button on the ribbon of the Veeam Backup browser to create an extra mount point on your VBR console server. From here you can copy the files to the restore file share in the protected environment a lot faster as well. All you need for this is a local user (non-administrator account) with the Veeam Restore Operator role.
Veeam Enterprise Manager addresses the challenges we worked around here. But you’ll need Enterprise (plus) to get full-featured use of that. It is nice as it takes care of the above scenario and you can have restore rights assigned to people without compromising your backup environments operational security. So, yes, use Enterprise Manager when possible. If not, see the above workarounds for other options.
Having some insights into how Veeam Backup & Replication works can be in handy at times. It also helps if you can think a bit outside the box and act on those insights to come up with some alternatives or workarounds. This is exactly what we did here with great results. The DBA could do his restores faster while the devs have the version of the database they needed. Do note that the correct answer lies in Veeam Enterprise manager, but lacking that. You now have some options. I hope these insights into a File-Level Restore in a hardened environment help you out someday.