Will regular SAMBA sharing work with ZFS filesystem? Do we need to create new mount points for that to work? I noticed that ZFS has its own SAMBA sharing that it did not work like I wanted.
zfs create mypool/KVM
zfs create mypool/NAS
zfs create mypool/NAS/Users
zfs create mypool/NAS/Media
zfs create mypool/System
mkdir mypool/NAS/Users/UserA
mkdir mypool/NAS/Users/UserB
mkdir mypool/NAS/Users/UserC
mkdir mypool/NAS/Media/Kids
mkdir mypool/NAS/Media/GeneralSo after doing some research, it seems like the spare motherboard I have doesn't support ECC RAM, and a max of 4GB at that.
Maybe QNAP it is . . .
RAM write-back.
Every modern OS should do similar things.
For example, you may have noticed that transferring files to a USB stick on Windows goes very fast at first, then slows down to the actual speed of the device. And when it's finished you still see the USB-stick flashing because its still writing data. This is the effect of write buffering.
This is because the application does not write to disk! The application uses API's provided by the operating system, which are handled by the filesystem and the virtual memory subsystem. Data read is cached, data written is buffered.
You are experiencing write buffering and when doing reads will experience file cache. Benchmarking is trickiy because before you know it, you are benchmarking your RAM and not your disk.
Can i see the dd-command that you used? And to what controller did you connect the disks? A RAID controller can do its own write-back. And if that is the case, you should know that a controller doing write-back can destroy the whole ZFS pool, no matter if you have a million mirror disks. Redundancy is not going to help you. Pretty much the only way to kill ZFS is to use a write-back controller that changes the order of writes across FLUSH request, which is a violation of the ATA-ACS2 specification. Not every RAID controller does this, but the more full-features ones do.
With 1TB you are pretty good. The general rule of thumb is at least 8 times your RAM size.
If you test with lower sizes however, you will notice the speed is higher. This is because part of the score is basically RAM-performance. The smaller your size is, the more percentage of that data will be caught by write-back. So you need a very large number to approach >99% where only minuscule amount is RAM-writeback.
That probably is an illusion. You test single throughput. Sure, this will give you comparable numbers. But on BSD platform the memory management of ZFS is much more mature, and difference in reality will occur. In many ways. I follow the commits of BSD HEAD development branch, and often see commits with performance increase in specific circumstances. Meaning, it is too simple to put performance in a single digit.In the end it looks like the performance is the same on both FreeNAS and Ubuntu.
Even if ZFS on Linux is less mature and less 'good' than on FreeBSD - which is true in my opinion - then it might still be good enough for you to use it, because you're more comfortable with Linux as platform for the things you wish to do with it.Now the big question is which one to go for! Both support all the features that I am looking for. Since this is one time setup and not something which I can keep changing like my desktop, I need to make an educated decision! FreeNAS is much easier to setup and use, but Ubuntu is completely configurable, and I am more comfortable with it compared to FreeBSD.
Meh, ecc ram is not absolutely required. The 4gb would be the bigger problem. Put the stuf together in a box and test out freenas or nas4free on it. See if you even like how it works first.
Why RAID5 stops working in 2009.
This is exactly one of the dataloss issues that ZFS will counter.
32 megabytes is not all that significant. And in all configurations ZFS uses less memory than virtually all other filesystems, including NTFS, FAT32 and Ext4.
The difference is that ZFS can use memory very effectively because it is smarter than the VFS-layer of the operating system, because it distinguishes between recent data (MRU) and popular data (MFU) which normal VFS does not and will throw away valuable caches agnostically.
ZFS will work fine with less RAM, but it won't be able to utilise all the speed potential of your harddrives. So with few RAM, your array might do 300MB/s instead of 700MB/s. That is the effect of more RAM for ZFS. Generally not a problem because gigabit will cap the speed anyway. More RAM can cache metadata which is very effective for file searching etc.
Most ZFS builds i help people with are quite cheap low-end builds of 200 - 300 for the system excluding harddrives. You can even do about 100-150 if your budget is stretched. This is cheaper than a solution from the shelf, which generally is twice as expensive while you get slower hardware and legacy software.
It will only protect against memory corruption, which is rather rare. Without ECC, ZFS can detect all and can correct at least some of the corruption caused by RAM bitflips. If it couldn't fix it, at least you know which files are affected. Also note that - even with ECC - you are still not protected against 3 bitflips; only against 1 (correction) and 2 (detection).
OpenZFS website had a feature-comparison i believe. But generally, the pool features do not vary much per platform. ZFS on Linux has pretty much the same on-disk features. That doesn't mean it supports the exclusive BSD features though; like automatic SWAP ZVOLs, booting from RAID-Z, TRIM on ZFS and more.How do you determine version/release number of ZFS on ubuntu and FreeNAS? I am trying to find out how far behind is ZFS on Linux compared to FreeBSD.
It doesn't? That would surprise me. SMART was one of the first features to make it into ZFSguru.FreeNAS has no opton to display SMART outputs.
Because it is very relevant; and exposes an inherent weakness of legacy RAID and filesystems.People are *still* quoting that article?
Well that is what the manufacturer tells you. And virtually all ZFS users i know have had uBER bad sectors on their consumer-grade drives already. So it's not some theoretical thing this happens all the time.Why? The assumptions are wrong. At the very least a raid controller that refuses to rebuild anything else because of a single bit error is downright malicious, and whether 1e-14 is an accurate number is certainly not established.