- #OPENZFS SCRUB ALL POOLS CODE#
- #OPENZFS SCRUB ALL POOLS PLUS#
Kicking off a scrub every day, week, or month doesn’t directly express that need. Scrub in its current form can take a very long time and can have highly deleterious impacts to overall performance.ĭata validation in ZFS should be specified according to data or business needs. Problem: ZFS does a great job detecting data errors due to lost writes, media, errors, storage bugs, but only when the user actually accesses the data.
property(?) to charge quota usage by before-compression-dedup size. You get 2G of quota, and any compression saved is to Google's benefit. Similar to how Google/Amazon/Cloud-Feature.acme already handles it. It is desirable as a provider, to be able to reverse this logic such that the customer's quota is reduced byĥ0MB and the 25MB compression/dedup saved, is to the provider's benefit. The customer's quota is only reduced by 25MB. ZFS Compression / Dedup to favour providerĬurrently, as a storage provider, if a customer has 100MB of quota available, and upload 50MB of data will reduce DDT entries from 512-bytes to 320-bytes. Use dedicated kmem_cache for deduplication table entries: RFC: zpool import -t for temporary pool names ()Ĭonvert synchronous writes to asynchronous writes when an ARC miss occurs during a lookup against the DDT. Improved SPA namespace collision management Subproject: document useful kernel interfaces for page manipulation on various platforms Replace larger ZIO caches with explicit pages Preferably compatible with pool version 29 for Solaris 10u11 compatibility. Specifying ashift when creating vdevs () The two sides of reflink() Unified ashift handling 
Recursive rollback (revert to a snapshot on dataset and all children, needs a new command line flag, -r is already taken)īased on indirect vdevs, rather than bprewrite.General feature ideas ZFS channel programs images for installing to real hardware.images could be built for running on public clouds.To easily try OpenZFS on a choice of distributions within a virtual machine: This would require a rewrite to not use Linux interfaces. Import ZFS on Linux sa=xattr into illumos.Įnsure that future additions/changes to the interface maintain maximum compatibility with userland tools.Įnable FreeBSD Linux jails / illumos lx brandz to use ZFS on Linux utilities. be able to run most of testrunner (and/or STF) test suite against userland implementation.be able to run /sbin/zfs, /sbin/zpool against userland implementation.expand this to also be able to test more of zfs in userland.We already have ztest / libzpool and want to: One sourcebase, rather than porting STF to every platform?
#OPENZFS SCRUB ALL POOLS CODE#
Make it easier to build, test, code review, and integrate ZFS changes into illumos. Mechanism for pull changes from one place to another Ideas for projects that would help coordinate changes between platforms …
#OPENZFS SCRUB ALL POOLS PLUS#
immediately pursuable ideas, plus long-term and strategic thoughts. Notes from the meeting that preceded Delphix's semi-annual Engineering Kick Off (EKO) 6.2.1 ZFS and OpenZFS in three minutes (or less)Īctive projects Notes from meetings Brainstorm, 18th September 2013. 4.9 ZFS Compression / Dedup to favour provider. 4.7 Improved SPA namespace collision management. 4.6 Replace larger ZIO caches with explicit pages. 3.7 Virtual machine images with OpenZFS. 3.6 Port ZPIOS from ZFS on Linux to illumos. 
3.5 /dev/zfs ioctl interface versioning. 3.1 Mechanism for pull changes from one place to another. The script is part of the Ubuntu package discussed here ĭon’t forget permissions : chmod 755 /opt/scrub. It should happen now and again, but if everything is going smooth sailing we tend to forget to do this With the ever increasing amount of ZFS servers around, it tends to be a hassle, so this is a perfect candidate for automation ! Details on the scrubbing 🙂Īwk 'BEGIN ' | \ So scrubbing ZFS: Giving ZFS a moment to sit down and think about its behavior.
0 kommentar(er)
