Just as we have over the last four or five years started losing the Internet/Computer pioneers, it's good to know their shoes are being filled people like these guys who will also go down in Computing history as pioneers. A beautiful technology.
I created ZFS-on-root on a 128GB flash drive - acts like any other drive using that filesystem. Decent on reads, fine if in ARC, decent as a native filesystem if accessing.
Been using it since 2008 in my home nas. At work NTFS it's not yet capable of any of this, quite embarrassing that most of the industry is stuck with obsolete solutions (for decades)
At 28 minutes in emphasis is put on how the checksum allows for self healing the data. However, there are checksums on the tree itself but no mention of what is done if the tree itself is bad. Seems like this situation must have been thought of before and has a valid backup solution too?
An average non-techy needs this to be given in the simpler terms without throwing abbreviations and terms left and right. This is if you want to bring in more enthusiastic neophytes into the realm of ZFS.
ZFS was sold for the enterprise. And you add capacity by plugging in a disk shelf with 48 drives at a time. Home usage and NAS use case where people just need a single drive came a long time after it went open source. It was made for the datacenter and that's why.
Just as we have over the last four or five years started losing the Internet/Computer pioneers, it's good to know their shoes are being filled people like these guys who will also go down in Computing history as pioneers. A beautiful technology.
Respect to Sun's engineers and the company mindset.
RIP Sun...
Thank you so much for posting this! This is gold!
So this is how block storage was invented. Amazing that Sun engineers thought about this when the cloud was not even born.
This is really helpful. Thanks so much, guys!
I created ZFS-on-root on a 128GB flash drive - acts like any other drive using that filesystem. Decent on reads, fine if in ARC, decent as a native filesystem if accessing.
Been using it since 2008 in my home nas. At work NTFS it's not yet capable of any of this, quite embarrassing that most of the industry is stuck with obsolete solutions (for decades)
At 28 minutes in emphasis is put on how the checksum allows for self healing the data. However, there are checksums on the tree itself but no mention of what is done if the tree itself is bad. Seems like this situation must have been thought of before and has a valid backup solution too?
Thanks for the video 😊
Was an OpenZFS video done for Scale 2019, with what's new?
Just for curiosity, is there also BTRFS talk mentioned in beginning?
An average non-techy needs this to be given in the simpler terms without throwing abbreviations and terms left and right.
This is if you want to bring in more enthusiastic neophytes into the realm of ZFS.
Is he using a Mac for the presentation? LOL
Still don't get why adding a drive to vdev was never a priority
ZFS was sold for the enterprise. And you add capacity by plugging in a disk shelf with 48 drives at a time. Home usage and NAS use case where people just need a single drive came a long time after it went open source. It was made for the datacenter and that's why.
I do not know nothing about zfs...
is it possible to build a Hardware OpenZFS raid Controller?
Can you stick 256 GB cache on a controller?
The beauty of this presentation is that he’s using a Mac with HFS or APFS. Go figure
Is that critical data we are talking about here?!...
If that's the only thing you have to say about that presentation, you are just a troll!
Irrelevant