10/27/2021 11:20:31 AM Error Installing Veeam Data Mover service Error: Failed to invoke command /opt/veeam/transport/veeamtransport --install 6162: /opt/veeam/transport/veeamtransport: error while loading shared libraries: libacl.so.1: cannot open shared object file: No such file or directory
Failed to invoke command /opt/veeam/transport/veeamtransport --install 6162: /opt/veeam/transport/veeamtransport: error while loading shared libraries: libacl.so.1: cannot open shared object file: No such file or directory
I use a Synology NAS as a off-site target in a GFS Copy Job.
It is a small installation, Hyper-V, 6-8 VMs, the backup server has local disks for the backup job and the off-site Synology as a iSCSI target formated with ReFS.
This has worked flawlessly for 2-3 years.
The problem with this setup is that the off-site target is mounted on the veeam server and is very exposed for ransomware if the veeam server is compromised.
I'm now wondering if it would be a idea to deploy a Linux VM on the Synology, and then use that VM as a Veeam hardened Linux repository.
Have anyone tried that?
Another idea would be to use a script to attach and detach the iSCSI repo before the copy job starts and after the copy job ends. That way the drive is only attached/visible to ransomware during the copy job.
Also, you are attaching an iSCSI repo over a WAN/off-site? Performance and latency don't kill you? I would think having a separate host at the off-site location with the iSCSI repo mounted would be a better option.
This has been an eye opening thread for me; thanks to all the contributors!
I'm looking to expand my vmware cluster storage with a Synology device; specifically for one file server that's taking up a lot of vsan space. This will *not* be a veeam repo.
Now I'm wondering: would there be any Veeam / Synology issues for this scenario: a Windows 2019 server vm on the cluster, using an iscsi connected synlology device as storage, and backing up to our current Veeam repository (a physical Supermicro box with Windows 2019, raid6 / ReFS)?
That is kinda a different topic, but to my understanding, Veeam wouldn't even try to back up the iSCSI drive because it isn't a drive in VMware. You CAN back it up if you use the Veeam agent, though.
If you're talking about having the NAS as part of VMware storage with the server on top of it, then my guess would be the same potential issues apply as the rest of the thread. The root cause of the issue is with the devices themselves, not how they are presented
Sorry for resurrecting an old thread but I'd really like to understand this issue better as I run a relatively new MSP and design Veeam backup systems for my small business customers. Let me know if I need to repost this as a new thread.
I use Synology NASes extensively for my repositories as well as some file servers. What would a Windows file server (with software/chipset raid or hardware raid) give me that the Synology does not? From reading y'alls prior discussion it looks like there are two issues at play:
-BBWC: from my understanding, this only protects against a sudden power failure or hard reboot of the NAS. As I have all of my NASes on UPSes with communication to trigger safe shutdowns, I feel that the risk of not having a BBWC is largely mitigated (although I'll grant not entirely).
-Protection against bit rot or other data corruption:
This one leaves me with a lot of questions. I know older versions of DSM ran on ext4 but newer versions offer BTRFS which enables checksum protection (and I always use BTRFS on new systems). Synology also offers periodic RAID scrubbing which, if I understand this correctly, checks the BTRFS checksums and RAID consistency between member disks. What data corruption safeguards does enterprise hardware raid controllers offer that BTRFS with RAID scrubbing does not? I want to understand if the recommendation against Synology NASes is based on the use of older technologies and may not apply to newer NASes with more modern file systems.
On a side note, I run most backup NASes with RAID 1 as my repositories are usually less than 16TB in size. Other than write performance, is there a case for RAID 5 over RAID 1 from a data corruption perspective since both offer single disk fault tolerance?
The questions around using Synology NAS devices as a file server are probably for another forum than Veeam. This particular post is around use of Synology NAS as a backup repo target. It's not really factoring in the source, which in your case might be data residing on another Synology NAS.
The bottom line on this post is integrating a Synology NAS into Veeam as a "standard Linux server over SSH" was sketchy at best. I created a script that would make it worked, and for a long time it did. But it no longer works with v12 and neither Veeam nor Synology believe this to be worth fixing. Synology has their own backup product they would rather push. And Veeam relies on the vendors to create integrations and such. Synology has effectively modded the OS too much for it to be considered a "standard Linux OS" nowadays.
This is why I have pivoted our deployments to using Synology repo by NFS. This unfortunately requires another infra component to mount the NFS, but it works. Only in a few very rare edge cases did I even have to resort to configuring virtualization on the Synology w/ a Linux VM to function in this capacity (situation was a remote offsite target with no other reasonable candidate or resources available).
This is not really a technical response to your post, but I think most of your technical questions are about Synology specifically and how it can or should or shouldn't be used in a production environment. Personally, as the senior backup architect for a large MSP, Synology is great for a backup target... but as you can see there are caveats even to that. We have not and would not recommend Synology to hold primary storage as we recommend Dell servers for that. But at the same time, I get that you're a new MSP and servicing small customers, so even a Synology NAS with software RAID10 would be better than a "powerful gaming workstation" that someone thought would be a good server.
Where performance isn't a major concern, as Synology NAS devices tend to not have the best comparable performance unless you get the most expensive rack mounted models, they're not bad as a backup destination. I previously managed backups at an MSP that exclusively used Synology devices for backup storage, also general file servers if the customer desired it. There are definite benefits, especially if you're managing mostly small businesses that likely have more limited IT budgets and don't want to get a closet full of servers for their business.
Synology's cloud-based DSM "QuickConnect" feature allows for remote management without network modifications, no need to open ports in a firewall or configure static IP addresses, which was one of the big benefits we saw as it virtually guaranteed that the Synology NAS would always be remotely accessible regardless of anything else that may get changed without someone knowing the consequences.
Also to my understanding the newer devices, which goes back at least a few years now, with BTRFS support should provide comparable data protection, against bitrot and such, with a configured RAID array. And they're definitely more affordable than a Windows server with comparable hardware, due mostly to Windows licensing costs.
That said, looking only at redundancy and fault-protection capabilities, you're probably safe with using the Synology NAS servers with BTRFS and RAID 1 as you mentioned.
Especially for selling to a small business, the Synology NAS was always an easier sell due to the reduced costs over a Windows server.
But looking at performance, either of a proper Linux server or a Windows server would likely be capable of higher speeds, though it might cost more depending on where the bottle neck is, network, drives, processing speed.
That's probably not the technical explanation you were looking for, but my overall recommendation is that the Synology servers are probably just fine and easier to sell to your customers due to costs, though you'd know your customers better than me. However, for long-term scalability at any customers that may need servers for other purposes later on, it may be worth selling them proper Windows servers due to future functionality, I wouldn't recommend using Synology servers to host an Active Directory environment or many other applications for instance, although most allow you to configure a virtual machine with whatever OS to serve whatever purpose you want if needed.
Side note, if you're using Windows Server for your backup repository and you're using ReFS, in order to properly take advantage of all that ReFS can do my understanding is that it's best to just use a JBOD configuration, leaving Windows to manage any redundancy and fault-protection. I'll admit though to not using ReFS presently myself so my experience with recent updates to ReFS is basically non-existent. Anyone feel free to correct me if I'm wrong on that.
We use Synology devices for back-up repo all the time for our customers.
We only use the Synology NAS as a host for block volumes (iSCSI).
We avoid using SMB and have never tried the old "Linux Server over SSH" trick.
You can do ReFS on a block volume presented through iSCSI from a Synology NAS to a Windows Server. But, it is not supported by Microsoft.
I have done this, but I suggest to only do it for back-up Copy repo, not for primary back-up repo.
If I would have to design a new architecture, I would probably try XFS instead of ReFS.
The supported way to use ReFS is (as @eSiloTim says) by using Windows Storage Spaces with JBOD or by using a hardware RAID conroller that is on the Hardware Compatibility List.
