34

One of the servers that I administrate runs the type of configuration that you describe. It has six 1TB hard drives with a LUKS-encrypted RAIDZ pool on it. I also have two 3TB hard drives in a LUKS-encrypted ZFS mirror that are swapped out every week to be taken off-site. The server has been using this configuration for about three years, and I've never had ...


33

I generally use one of the following two filesystems: XFS for anything which does not play well with CoW (or for virtual machines whose datastore already is on a CoW filesystem) or when extremely fast direct I/O is required; ZFS for anything else. For your use case I would use ZFS, especially considering that Ubuntu 18.04 already ships it. As you can ...


13

BINGO!!! Alright, here is how this was resolved: I referenced this artice for unmounting the datastore, https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2004605 Step 1: Manually unmount the datastore from the host with this command: esxcli storage filesystem unmount [-u UUID | -l label | -p path ] In ...


9

NET USE does not allow you to connect to the same server with different credentials. This is because it reuses the same relationship to map additional shares on the same server. You can work around this by making windows think it's a different server either by creating DNS aliases for the same NAS with different names or editing the hosts file on the clients....


9

rsync doesn't do SFTP. From the man page: There are two different ways for rsync to contact a remote system: using a remote-shell program as the transport (such as ssh or rsh) or contacting an rsync daemon directly via TCP. SFTP doesn't give you a shell, ergo it doesn't work for rsync. You'll need an SSH connection instead. Once you've ...


8

Synology has an App on it's DSM store called Storage Analyzer, which provides a very friendly user interface to view usage by folder, user, file types, duplicates, etc... It also keeps a history by analyzing periodically...


8

Most NAS snapshots implement copy-on-write, so the snapshot itself initially takes up no space (or next-to-none, there is some overhead). But any block that is changed whilst the file it's part of is snapshotted must be copied. If an entire file is deleted whilst it's still in a snapshot, all those blocks must still be kept. So in your case above, as long ...


8

LACP always helps with redundancy but isn't your panacea for performance. https://thenetworkway.wordpress.com/2015/05/01/an-overview-of-link-aggregation-and-lacp/ TL;DR: You really need multiple TCP connections to get performance boost, and this isn't what's always possible... SMB can do that "automagically" so you either LACP or SMB Multichannel but say ...


8

The NAS is a 2U Synology Rack set up in SMB 2.0 and higher mode. Plenty of space on the share (and NAS). The server is Windows 2012 R2 SMB 2.xx is your problem. You need SMB 3.xx to allow Hyper-V running VMs from your file share. https://blogs.technet.microsoft.com/yungchou/2012/09/10/windows-server-2012-hyper-v-over-smb-explained/


8

Nowadays the border between NAS and SAN is unclear especially for SMB products. Historically NAS devices consist of the bunch of disks arranged in some RAID to maximize the overall performance and achieve the required data redundancy. These devices have core operating systems aboard. Those OS do all low-level processing and file systems maintenance. File ...


7

We are using Veeam generally over the whole company's infrastructure and it works great for a years. AFAIR you can even use free version. https://www.veeam.com/free-virtualization-software-vmware-hyper-v.html


7

SAN/NAS for < $3,600 per TB? This is totally doable. However, don't forget that apart from SAN hardware/software itself, there are lots of things that should be considered, such as: redundant power supplies, redundant WAN/LAN, IT staff onsite who will manage your infrastructure, etc.. First off, define how much downtime costs for you company and ...


6

Definitely agree with both answers already provided here. If head office is running on a single host you should probably get a second one regardless of storage latency - phoebus Nov 1 '13 at 12:17 My understanding is that average latency for a VMFS datastore should be 20ms or less, so I don't see a problem here. – joeqwerty Nov 1 '13 at 15:12 I would dig ...


6

Yes, it's still the convention, and yes, it holds true even as you scale. With ZFS, in fact, you really don't want to get to the 75% mark in your zpool too often. Fragmentation, snapshots and general performance tend to be impacted. If building anew, don't start with anything more than 40% utilization and be sure to plan for growth.


6

This tip worked for me in the past, if you can disable ATS-Only: Cannot mount the VMFS5 datastore when using storage encryption appliances (2030416) I guess no VM are up in your case, but they need to be close to be able to run the command. Some storage encryption appliances may block certain SCSI commands, including ATS commands, sent from ESXi ...


6

RAID 10 is just stripes over mirrors. That means for every mirror pair only one drive may fail. You also lose 50% of the storage. I don't know how your NAS handles two RAID 1 arrays but it might be a bit faster to use RAID 10. If you want better fault tolerance you could use RAID 6. There you still have 50% capacity loss but any two drives may fail. RAID ...


6

I have reproduced this scenario in my test environment and I could find my data normally. To reproduce it I've followed these steps. Make sure that you follow every steps. Editing this file needs root access; therefore you will need to use sudo with your command. You can also open the file in any of your personal favorite text editors. 1 - Installed and ...


6

This looks more like a question for superuser than for serverfault, but some of the ideas are valid for this site too, so I'll take a stab at answering some of the questions: XFS has had a reputation of not liking power loss. A lot has happened since, and RedHat, Oracle and the like use it as a default file system nowadays, so I wouldn't be surprised if ...


6

They use MDRAID+LVM and they use their proprietary software RAID implementation. Second one should be avoided due to smaller user base and recovery issues. https://www.synology.com/en-us/knowledgebase/DSM/tutorial/Storage/What_is_Synology_Hybrid_RAID_SHR


5

That's bittorrent traffic. More specifically, it's traffic caused by the distributed hash table (DHT) protocol. There's even a thread about it on the QNAP forums


5

For some background: The reason you typically get a NAS appliance is to get something that will serve up storage and "just work". Of course you need to do an initial configuration, but after that, you pretty much just leave it alone until a drive fails, or there's a critical firmware update. For a Windows file server, you'll need to do regular ...


5

Storage latency will be your metric. If your application is highly sensitive to storage latency you'll want to steer clear of AWS and go physical, or drop the money to get the Storage Optimized instances. They specifically state those are the types for things like Hadoop and Cassandra. The thing about the higher tiers of AWS instance types is that it isn't ...


5

quota resize (or the equivalent action in one of the GUIs) forces the quota service to scan the volume and apply any changes made to /etc/quotas. It's not actually resizing your volume. Until you execute a quota resize (or alternatively disable and reenable the quota service), no changes made to the quota definitions will be applied. Also notable: during the ...


5

D. None of the above. Network link saturation is rarely the cause of performance problems, in my experience. Additionally, it sounds like you're trying to solve a problem that you don't even know exists, which is a bit like tilting at windmills. That model has 1Gbe Ethernet ports, I'm skeptical of the idea that a single 1Gbe link is going to be saturated ...


5

As noted you will definetely need at least one more NAS to build truly HA environment. Then you connect them as DAS to each of the host, create shared storage across and configure Failover cluster. From that point, you' ll be able to go with SQL AAQ feature, for example. To make a storage pool, check some SDS solutions like HP VSE or StarWind. Not sure about ...


5

It's not going to work reliably. NTFS (or ReFS) aren't cluster-aware file systems, multi-mount is going to destroy metadata tables immediately. Either you use SMB3 bolt-on with an arbitration thing on top (that's what Microsoft calls Cluster Shared Volume or CSVFS), or you use third-party arbiter like MetaSAN. Good story could be found here: Trying to be ...


5

Yes, you can do this while online. Data or metadata references are only updated once a balance has completed for a particular chunk, so it will remain consistent even during modification. If the system is highly write transactional, the balance operation will take a bit longer, but that's a lot better than having to take the system offline.


5

Look at StarWind VSAN 2-nodes hyperconverged solution on VMware here. It gives a possibility to mirror storage between the hosts and present it as HA datastore to the cluster. Virtual Machines can be deployed then on a top of HA datastore. Additionally, they have a ready for Production hardware solution - StarWind HyperConverged Appliance.


5

Just want to address a few of your points. Full disclosure, I am a proponent of ZFS. This is all IMHO. RE: your needs: It should not need constant tinkering and maintenance. As long as all your scrubs, snapshots, security updates, capacity alerts, etc. are configured correctly, it should just work (at least until it phones home). Buggy software (...


5

I would use whatever journalling file system the operating system in question suggested as default unless I had very good reason not to. Last time I checked with Ubuntu that was ext4. The reason is simple: The default is the one that most likely is used the most, so the chance for any bugs to have been found and fixed is the largest. The differences ...


Only top voted, non community-wiki answers of a minimum length are eligible