Posts by ellnic

    This is not something I have experienced, but give me some time today and I will see if I can reproduce.

    Well IDK who Jody Bruchon is... nor does anyone else. He's certainly no one the industry recognises as a file system expert - or has even heard of for that matter. He looks like some guy with his own, a blog, and he likes youtube.. and a bad opinion on everything. Don't forget the awesome films he's directed. It was quite comical quoting such an awful reference.. I think it made the point perfectly.

    This one I don't understand.
    Why should this happen with Ext4?

    What you need here is to understand bit rot.

    I said to this other person outside of this forum too, that ZFS is not some kind of magic which is capable of everything.

    Really? Damn.. I was going to ask it to predict this weeks lottery numbers.

    And even if photos are gone: every person should have backups around.

    Of course, and ZFS helps mitigate the risk of those backups being junk.

    The other person outside reacted exactly like this too. I never said ZFS users are stupid.

    .... I don’t think this can be explained to you.

    Whatever. I'm back to Ext4, all my data is safe with that too.

    Then that’s all that matters, isn’t it? If you are happy that your data is safe, then there’s no need to change anything. To quote the Inbetweeners: “You can lead a horse to water, but you can’t make it stick Lego up it’s bum.”

    It's working but on a 1 TB drive i Only can access 898 GiB. This is way too loosy.

    Have you set the correct ashift for your drive?

    I will get away from ZFS. Furthermore I will get away because all persons using ZFS I know totally stress me out using datasets, Snapshots, RAID and all this stuff I absolutely don't need.
    I use rsync once a day or a week and I have my data always safe on an external drive. Snapshots and all this stuff are totally stupid and only for people with too much time.

    These people just thing ZFS is the wonder of IT and better than everything else. Nothing can damage your data. They even say zfs send <whatever> sends 5 TB in one hour. I don't like liars. So I will stay away from that.

    I go back to EXT4.

    Fine, you do that. No one is forcing you to use it. You came here asking for help getting it up and running, but what you really have a problem with is the implementation of the filesystem itself. You go back to EXT4. I'll stick with my ZFS datasets, so when I open my family photos that I haven't looked at for a while, they're not half gone and my music remains blip free. I don't think I am stupid for wanting to use ZFS - what I do think is stupid is coming to a forum where other people are helping you in their free time and then throwing your toys out the cot.

    This guy always talked about meta data and pointers who get saved.
    Snapshots are fine, but not the zfs ones.

    I use rsync to my second and third disk and if someone deltes something, I can get it back easily too, no problem :P No over-complicated zfs needed :P

    What you are talking about is normal backup. Backup which will backup a corruption if it occurs. But ultimately it's up to you what you choose to use. If you are happy with what you are using, then by all means carry on that way - we stupid ZFS users won't stop you.

    I am a normal user, no pro. All I want is that it's working without being in ssh 10 hours a day.

    I don't SSH 10 hours a day, and OMV makes this so simple you don't need to. If you're referring to the commands I pointed you to earlier, they were just to help you figure out if compression was worthwhile and they wouldn't be done more than once.

    I'm pretty sure 7zip is lz4 which is what you would be best choosing as a compression method for ZFS if you go down that route - so a quick guess, no. But you'd be better off trying it on your WD green and checking. ZFS can tell you how much space you are saving.

    This is from one of my media pools:

    compressratio 1.01x

    As you would expect, there is not much saving. But you will find that changes depending on the files you put in the pool. LZ4 can give you very good savings with VERY little CPU cost. It's a bit of a miracle algorithm.

    Once you've setup your pool, turn compression on:

    zfs set compression=lz4 [Poolname]

    Copy data in, then:

    zfs get all [Poolname] | grep compressratio

    I will test it out with my 1 GB WD Green soon. It's a trash drive so, perfect for testing.

    This would be better.

    Will the ZFS plugin be supported a long time in OMV? Don't want to set up ZFS and then in a few updates I can't use it anymore.

    I don't actually use the Plugin to create and administer my pools, I use the command line. I use the plugin because it integrates well with the other settings. I can't see support going anytime soon. A lot of work has been put into it of late. It's unlikely to just get abandoned.

    I got all excited then for a minute.. I thought Proxmox might offer some kind of livepatch, but their solution is to use Kernel Care ;-(

    This would be another win if OMV moved to Ubuntu. Canonical allow you to register 3 systems for free. That's probably one of the best automagic offerings going.

    Because it is the Ubuntu 18 LTS kernel, it might have the changes backported to it.

    I had a look for a changelog.. all I could find is this:;a=summary

    I'm somewhat surprised to find; with OMV4 - Kernel 4.17 and the ZFS 4.0.4 plugin, that a ZFS mirror is running well on an Atom D525 processor with 4GB RAM. That's certainly not fast hardware.
    I still have a bit of testing to do, but it looks pretty good so far.

    Just bare in mind that the D525 doesn't have ECC.

    Really? I misread that because I think I wanted to assume you were talking about Oracle.. I thought the almighty Red Hat would win on this... after all, the support they sell is basically their business. You’d think they’d want to shine above the likes of Oracle. If review sites are anything to be believed, they seem to put Oracle in favourable light.

    Ah ok. Yeah that makes sense then. I wish more distros would follow suit. If Canonical are doing it, the smaller outfits should follow.

    Oh I didn’t realise Oracle support was so bad. I generally avoid RH/Centos/OEL... I’ve revisited from time to time, and I always come back to Debian / Ubuntu.

    What I'd like to know is how they are managing to get away with that... I thought the Ubuntu installations just matched the Kernel with the ZFS files in the main repo... ergo, it 'comes' with ZFS after a single apt. They always make sure everything matches etc - so a simple install procedure. But what this seems to suggest with the Proxmox kernel is that ZFS is being bundled with a GPLv2 project... or am I mistaken? Kudos to them if it is.... I'm starting to think the reason Oracle aren't changing the license is because they don't have permission of all coders - but they don't actually care. They won't do it, because that would be seen as a really bad example, but they couldn't care less (or even applaud?) those who do? Who knows.... why haven't they taken legal action if there are these 'breaches'?

    Oh and while we are on the subject of dear old Oracle.... I'm gonna make some enemies here... Oracle Linux is better than Centos. Yep.. I said it. I have changed my opinion. Why? Upstream delay is a LOT less and they use an Ubuntu style 'pay if you want'. I'd normally steer clear of anything Oracle, but if you want RPM that's binary compatible to Redhat (but not) then OEL wins. *runs and hides*

    Edit: This is not to say that I think Red Hat (and derivatives) are better than Debian (and derivatives). Just that if I were forced to use a Red Hat distro with a zero budget that it wouldn't be Centos.