Return to zfs.mdwn CVS log

Up to [NetBSD Developer Wiki] / wikisrc

zfs: formatting

    1: # ZFS on NetBSD
    2: 
    3: This page attempts to do two things: provide enough orientation and
    4: pointers to standard ZFS documentation for NetBSD users who are new to
    5: ZFS, and to describe NetBSD-specific ZFS information.  It is
    6: emphatically not a tutorial or an introduction to ZFS.
    7: 
    8: Many things are marked with \todo because they need a better
    9: explanation, and some have question marks, indicating that the
   10: statement needs verification.
   11: 
   12: # Documentation Pointers
   13: 
   14: See the man pages for zfs(8) and zpool(8).
   15: 
   16:   - [Oracle ZFS Administration Manual](https://docs.oracle.com/cd/E26505_01/html/E37384/index.html)
   17:   - [FreeBSD Handbook ZFS Chapter](https://www.freebsd.org/doc/handbook/zfs.html)
   18:   - [OpenZFS admin docs index page](https://github.com/openzfs/zfs/wiki/Admin-Documentation)
   19: 
   20:   - [Wikipedia](https://en.wikipedia.org/wiki/ZFS)
   21: 
   22: # Status of ZFS in NetBSD
   23: 
   24: ## Sources of ZFS code
   25: 
   26: \todo Verify/fix.
   27: 
   28: Currently, there are multiple ZFS projects and codebases:
   29: 
   30:   - ZFS as released under the CDDL (common ancestor)
   31:   - [OpenZFS](http://www.open-zfs.org/wiki/Main_Page) [github wiki](https://github.com/openzfs/zfs/wiki)
   32:   - [zfsonlinux](https://zfsonlinux.org/)
   33:   - [OpenZFS on OS X ](https://openzfsonosx.org/) [repo](https://github.com/openzfsonosx)
   34:   - proprietary ZFS in Solaris (not relevant in open source)
   35: 
   36: OpenZFS is a coordinating project to align open ZFS codebases.  There
   37: is a notion of a shared core codebase and OS-specific adaptation code.
   38: As of early 2020, it appears that there are two co-root repositories:
   39: illumos and zfsonlinux, with a notion that fixes and improvements
   40: should be cross-shared.  This is different from the situation through
   41: late 2019 where sharing was done via a designated sharing repository,
   42: and appears to be a new, more efficient, procedure among cooperating
   43: people, rather than any kind of fork.
   44: 
   45:   - [deprecation of OpenZFS repo](https://github.com/openzfs/openzfs/)
   46:   - [zfsonlinux relationship to OpenZFS](https://github.com/openzfs/zfs/wiki/OpenZFS-Patches)
   47: 
   48: \todo Explain how FreeBSD code relates to zfsonlinux (imported/merged and fixes pushed upstream?).
   49: 
   50: \todo Explain how the NetBSD code relates (imported from FreeBSD? intent to continue to track?)
   51: 
   52: \todo Explain if NetBSD has a plan to join zfsonlinux as a first-class
   53: member, vs tracking via FreeBSD.
   54: 
   55: \todo Explain how OpenZFS on OS X relates (tracks zfsonlinux?).
   56: 
   57: See [FreeBSD's history](https://wiki.freebsd.org/ZFSTuningGuide).
   58: 
   59: ## NetBSD code history
   60: 
   61: \todo This section really needs help.
   62: 
   63: Before NetBSD 8, NetBSD imported ZFS code from ?, which was version ?
   64: 
   65: Before NetBSD 9, NetBSD imported updated ZFS code from FreeBSD.  That
   66: FreeBSD code came from ?, which was version ?
   67: 
   68: ## NetBSD 8 and earlier
   69: 
   70: While there is some ZFS code, it is old, and seems to have significant
   71: problems.  If one wants to use ZFS, first upgrade to NetBSD 9.  It is
   72: unlikely that anyone is interested in helping, other than telling you
   73: to upgrade to 9.
   74: 
   75: (Reports of how well NetBSD 8 works are welcome on netbsd-users, if it
   76: can actually be recommended for use.)
   77: 
   78: ## NetBSD 9
   79: 
   80: There have been fixes since 9.0 RELEASE.  It is best to upgrade along
   81: the netbsd-9 branch, but the release should be ok.  Most aspects work solidly.
   82: 
   83: \todo Explain this in terms of versions of FreeBSD OpenZFS and/or
   84: zfsonlinux.
   85: 
   86: \todo This supports pool version 28/5000 (really true?).  Of the
   87: feature flags found in modern OpenZFS, \todo are supported.
   88: 
   89: Generally, fixes to ZFS in current will be pulled up to 9, but new
   90: features typically will not be.
   91: 
   92: ## NetBSD current
   93: 
   94: The ZFS code in current is very similar to that in 9.
   95: 
   96: There is initial support for [[ZFS root|wiki/RootOnZFS]], via booting from
   97: ffs and pivoting.
   98: 
   99: One can make a ccd using a zvol as a component.  See the zvol section below.
  100: 
  101: ## Things that aren't supported yet
  102: 
  103: \todo hotswap (maybe - not clear exactly what this means)
  104: 
  105: \todo direct boot into zfs root (via boot blocks reading zfs)
  106: 
  107: ## Architectures
  108: 
  109: Most people seem to be using amd64.
  110: 
  111: To build zfs, one puts MKZFS=yes in mk.conf.  This is default on amd64
  112: and aarch64 on netbsd-9.  In current, it is also default on sparc64.
  113: 
  114: More or less, zfs can be enabled on an architecture when it is known
  115: to build and run reliably.  (Of course, users are welcome to build it
  116: and report.)
  117: 
  118: # NetBSD-specific information
  119: 
  120: ## rc.conf
  121: 
  122: The main configuration is to put zfs=YES in rc.conf, so that the rc.d
  123: scripts bring up ZFS and mount ZFS file systems.
  124: 
  125: ## pool locations
  126: 
  127: One can add disks or parts of disks into pools.  Methods of specifying
  128: areas to be included include:
  129: 
  130:   - entire disks (e.g., /dev/rwd0d on amd64)
  131:   - disklabel partitions (e.g., /dev/sd0e)
  132:   - wedges (e.g., /dev/dk0)
  133: 
  134: ## legacy vs ? mount points
  135: 
  136: \todo Explain, if this is NetBSD specific.  Explain consequences, as
  137: this seems to have something to do with mount ordering.
  138: 
  139: ## mount order
  140: 
  141: NetBSD 9 mounts other file systems and then ZFS file systems.  This can
  142: be a problem if /usr/pkgsrc is on ZFS and /usr/pkgsrc/distfiles is on
  143: NFS.  A workaround is to use noauto and do the mounts in
  144: /etc/rc.local.
  145: 
  146: NetBSD current after 20200301 mounts ZFS first. \todo Explain
  147: consequences.
  148: 
  149: ## NFS
  150: 
  151: \todo Verify if this is accurate.
  152: 
  153: zfs filesystems can be exported over NFS.  While there are zfs
  154: commands that appear to be about controlling exports, they simply
  155: print information that can be added to exports(5).
  156: 
  157: This is reported to work on 9.0 STABLE, but to cause a panic on
  158: current (20200302).
  159: 
  160: ## zvol
  161: 
  162: Within a ZFS pool, the standard approach is to have file systems, but
  163: one can also create a zvol, which is a block device of a certain size.
  164: 
  165: \todo The zvol will appear as /dev/???? and can be used in many
  166: respects like a slice.  However, the system will not read disklabels
  167: and gpt labels from a zvol; in this respect it is more like a disklabel
  168: partition or wedge than a disk drive.
  169: 
  170: \todo Explain that one can export a zvol via iscsi.
  171: 
  172: \todo Explain if one can swap on a zvol.
  173: 
  174: \todo Explain that one can use ccd to create a normal-looking disk
  175: from a zvol.  This allows reading a GPT label from the zvol, which is
  176: useful in case the zvol had been exported via iscsi and some other
  177: system created a label.
  178: 
  179: ## TRIM
  180: 
  181: There is some notion of TRIM and zfs using it.
  182: 
  183: \todo Explain how this relates to NetBSD.
  184: 
  185: # Memory usage
  186: 
  187: Basically, ZFS uses lots of memory and most people run it on systems
  188: with large amounts of memory.  NetBSD works well on systems with
  189: comparatively small amounts of memory.  So a natural question is how
  190: well ZFS works on one's VAX with 2M of RAM :-)
  191: 
  192: More seriously, one might ask if it is reasonable to run ZFS on a RPI3
  193: with 1G of RAM, or even if it is reasonable on a system with 4G.
  194: 
  195: \todo Give ballpark level for minimum sane RAM, and the amount which
  196: is cleanly enough.
  197: 
  198: For now, a good guess is that a 4G system with only 1T of disk is
  199: probably ok, and that 1G is very likely not ok.
  200: 
  201: FreeBSD has some documentation about memory use.  There is a notion of
  202: a minimum of 1G (used for ZFS), and using 1G for 1T of storage, and
  203: more if deduplication is enabled.  FreeBSD considers all i386 systems
  204: to be low memory; this appears to be a clue.
  205: 
  206: \todo Explain if the FreeBSD sysctl list applies, or if not what we
  207: should do instead.
  208: 
  209:   - [FreeBSD low memory documentation](https://www.freebsd.org/doc/handbook/zfs-advanced.html)
  210: 
  211: # Interoperability with other systems
  212: 
  213: \todo Explain pool version and feature flags relationship to FreeBSD,
  214: Linux, OpenIndiana/Illumos/?, and ?
  215: 
  216: \todo Explain how to configure a pool in terms of version/features for
  217: use with particula other systems.
  218: 
  219: # Quick Start
  220: 
  221: See the [FreeBSD Quickstart
  222: Guide](https://www.freebsd.org/doc/handbook/zfs-quickstart.html); only
  223: the first item is NetBSD specific.
  224: 
  225:   - Put zfs=YES in rc.conf.
  226: 
  227:   - Create a pool as "zpool create pool1 /dev/dk0".
  228: 
  229:   - df and see /pool1
  230: 
  231:   - Really, read the FreeBSD docs and the other linked documentation above.