rvassar
Guru
- Joined
- May 2, 2018
- Messages
- 972
ZFS fans, rejoice—RAIDz expansion will be a thing very soon
Founding OpenZFS dev Matthew Ahrens opened a pull request last week.
arstechnica.com
-
Important Announcement for the TrueNAS Community.
The TrueNAS Community has now been moved. This forum will now become READ-ONLY for historical purposes. Please feel free to join us on the new TrueNAS Community Forums
Oh my!!! ZFS Expansion in the news...
- Thread starter rvassar
- Start date
arstechnica.com
rvassar
Guru
- Joined
- May 2, 2018
- Messages
- 972
Amusingly..."Historically, ZFS has eschewed this sort of expansion. ZFS was originally developed for business use, and live array re-shaping is generally a non-starter in the business world. Dropping your storage's performance to unusable levels for days on end generally costs more in payroll and overhead than buying an entirely new set of hardware would."
For the most part, only enterprise RAID HW supports migration from one RAID mode to another...
For the most part, only enterprise RAID HW supports migration from one RAID mode to another...
- Joined
- Jan 1, 2016
- Messages
- 9,703
A few points about that...
RAIDZ expansion isn't a change in RAID mode. (you're still stuck with RAIDZn where n is whatever you started with... you can only vary the width of your VDEV... from 4 to 5 disks, for example)
The array isn't rebuilt with data redistributed like it is with proprietary RAID array controllers, rather new (empty) blocks held for existing data to make it look like N+1 (even though there are only N blocks in the set) and then all new data written as N+1. Up-side to that is no enormous performance hit for days to complete the change (although there is a lot of work done on metadata). Disclaimer: this is based on my understanding from a presentation I watched by Allan Jude some years ago... things could have changed.
I personally wouldn't want to get involved with that on my systems.
I see how the SOHO market comes to TrueNAS from QNAP or SYNOLOGY and expects it and often are very unhappy to find that (not having done their research in advance) it doesn't work that way.
There's no substitute for good design up-front... But I guess there are a few scenarios that you could find yourself in as a result of reality happening despite the best plans and this could provide an additional tool in the belt to work your way out of the situation.
I would not consider it good practice to start with a small pool and just keep growing it by one disk as needed over time... this would be wasting a lot of space by the end if nothing else.
RAIDZ expansion isn't a change in RAID mode. (you're still stuck with RAIDZn where n is whatever you started with... you can only vary the width of your VDEV... from 4 to 5 disks, for example)
The array isn't rebuilt with data redistributed like it is with proprietary RAID array controllers, rather new (empty) blocks held for existing data to make it look like N+1 (even though there are only N blocks in the set) and then all new data written as N+1. Up-side to that is no enormous performance hit for days to complete the change (although there is a lot of work done on metadata). Disclaimer: this is based on my understanding from a presentation I watched by Allan Jude some years ago... things could have changed.
I personally wouldn't want to get involved with that on my systems.
I see how the SOHO market comes to TrueNAS from QNAP or SYNOLOGY and expects it and often are very unhappy to find that (not having done their research in advance) it doesn't work that way.
There's no substitute for good design up-front... But I guess there are a few scenarios that you could find yourself in as a result of reality happening despite the best plans and this could provide an additional tool in the belt to work your way out of the situation.
I would not consider it good practice to start with a small pool and just keep growing it by one disk as needed over time... this would be wasting a lot of space by the end if nothing else.
