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: - [OpenZFS Documentation](https://openzfs.github.io/openzfs-docs/)
17: - [OpenZFS admin docs index page](https://github.com/openzfs/zfs/wiki/Admin-Documentation)
18: - [FreeBSD Handbook ZFS Chapter](https://www.freebsd.org/doc/handbook/zfs.html)
19: - [Oracle ZFS Administration Manual](https://docs.oracle.com/cd/E26505_01/html/E37384/index.html)
20: - [Wikipedia](https://en.wikipedia.org/wiki/ZFS)
21:
22: # Status of ZFS in NetBSD
23:
24: ## NetBSD 8
25:
26: NetBSD 8 has an old version of ZFS, and it is not recommended for use
27: at all. There is no evidence that anyone is interested in helping
28: with ZFS on 8. Those wishing to use ZFS on NetBSD 8 should therefore
29: update to NetBSD 9.
30:
31: ## NetBSD 9
32:
33: NetBSD-9 has ZFS that is considered to work well. There have been
34: fixes since 9.0_RELEASE. As always, people running NetBSD 9 are
35: likely best served by the most recent version of the netbsd-9 stable
36: branch. As of 2021-02, ZFS in the NetBSD 9.1 release is very close to
37: netbsd-9.
38:
39: ## NetBSD-current
40:
41: NetBSD-current (as of 2021-02) has similar ZFS code to 9.
42:
43: There is initial support for [[ZFS root|wiki/RootOnZFS]], via booting from
44: ffs and pivoting.
45:
46: ## NetBSD/xen special issues
47:
48: In NetBSD-9, MAXPHYS is 64KB in most places, but because of xbd(4) it
49: is set to 32KB for XEN kernels. Thus the standard zfs kernel modules
50: do not work under xen. In NetBSD-current, xbd(4) supports 64 KB
51: MAXPHYS and this is no longer an issue.
52:
53: Xen and zfs on current are reported to work well together, as of 2021-02.
54:
55: ## Architectures
56:
57: Most people seem to be using amd64.
58:
59: To build zfs, one puts MKZFS=yes in mk.conf. This is default on amd64
60: and aarch64 on netbsd-9. In current, it is also default on sparc64.
61:
62: More or less, zfs can be enabled on an architecture when it is known
63: to build and run reliably. (Of course, users are welcome to build it
64: and report.)
65:
66: # Quick Start
67:
68: See the [FreeBSD Quickstart
69: Guide](https://www.freebsd.org/doc/handbook/zfs-quickstart.html); only
70: the first item is NetBSD specific.
71:
72: - Put zfs=YES in rc.conf.
73:
74: - Create a pool as "zpool create pool1 /dev/dk0".
75:
76: - df and see /pool1
77:
78: - Create a filesystem mounted on /n0 as "zfs create -o
79: mountpoint=/n0 pool1/n0".
80:
81: - Go back and read the documentation and start over.
82:
83: # NetBSD-specific information
84:
85: ## rc.conf
86:
87: The main configuration is to put zfs=YES in rc.conf, so that the rc.d
88: scripts bring up ZFS and mount ZFS file systems.
89:
90: ## pool locations
91:
92: One can add disks or parts of disks into pools. Methods of specifying
93: areas to be included include:
94:
95: - entire disks (e.g., /dev/wd0d on amd64)
96: - disklabel partitions (e.g., /dev/sd0e)
97: - wedges (e.g., /dev/dk0)
98:
99: ## mount order
100:
101: NetBSD 9 mounts other file systems and then ZFS file systems. This can
102: be a problem if /usr/pkgsrc is on ZFS and /usr/pkgsrc/distfiles is on
103: NFS. A workaround is to use noauto and do the mounts in
104: /etc/rc.local.
105:
106: NetBSD current after 20200301 mounts ZFS first. The same issues and
107: workarounds apply in different circumstances.
108:
109: ## NFS
110:
111: zfs filesystems are reported to be exportable over NFS.
112:
113: The "zfs share" command adds a line to /etc/zfs/exports, and "zfs
114: unshare" removes it.
115:
116: \todo Explain how /etc/zfs/exports is used.
117:
118: This is reported to work on 9.0 STABLE, but to cause a panic on
119: current (20200302). See [misc/55042](http://gnats.netbsd.org/55042).
120:
121: ## zvol
122:
123: Within a ZFS pool, the standard approach is to have file systems, but
124: one can also create a zvol, which is a block device of a certain size.
125:
126: \todo The zvol will appear as /dev/???? and can be used in many
127: respects like a slice. However, the system will not read disklabels
128: and gpt labels from a zvol; in this respect it is more like a disklabel
129: partition or wedge than a disk drive.
130:
131: \todo Explain that one can export a zvol via iscsi.
132:
133: \todo Explain if one can swap on a zvol.
134:
135: \todo Explain that one can use ccd to create a normal-looking disk
136: from a zvol. This allows reading a GPT label from the zvol, which is
137: useful in case the zvol had been exported via iscsi and some other
138: system created a label.
139:
140: # Memory usage
141:
142: Basically, ZFS uses lots of memory and most people run it on systems
143: with large amounts of memory. NetBSD works well on systems with
144: comparatively small amounts of memory. So a natural question is how
145: well ZFS works on one's VAX with 2M of RAM :-) More seriously, one
146: might ask if it is reasonable to run ZFS on a RPI3 with 1G of RAM, or
147: if it is reasonable on a system with 4G.
148:
149: The prevailing wisdom is more or less that ZFS consumes 1G plus 1G per
150: 1T of disk. 32-bit architectures are viewed as too small to run ZFS.
151:
152: Besides RAM, zfs requires that architecture kernel stack size is at
153: least 12KB or more -- some operations cause stack overflow with 8KB
154: kernel stack. On NetBSD, the architectures with 16KB kernel stack are
155: amd64, sparc64, powerpc, and experimental ia64, hppa. mac68k and sh3
156: have 12KB kernel stack. All others use only 8KB stack, which is not
157: enough to run zfs.
158:
159: NetBSD has many statistics provided via sysctl; see "sysctl
160: kstat.zfs".
161:
162: FreeBSD has tunables that NetBSD does not seem to have, described in
163: [FreeBSD Handbook ZFS Advanced
164: section](https://docs.freebsd.org/en/books/handbook/zfs/#zfs-advanced).
165:
166: # Interoperability with other systems
167:
168: Modern ZFS uses pool version 5000 and feature flags.
169:
170: It is in general possible to export a pool and them import the pool on
171: some other system, as long as the other system supports all the used
172: features.
173:
174: \todo Explain how to do this and what is known to work.
175:
176: \todo Explain feature flags relationship to FreeBSD, Linux, iIllumos,
177: macOS.
178:
179: # Sources of ZFS code
180:
181: Currently, there are multiple ZFS projects and codebases:
182:
183: - [OpenZFS](http://www.open-zfs.org/wiki/Main_Page)
184: - [openzfs repository](https://github.com/openzfs/zfs}
185: - [zfsonlinux](https://zfsonlinux.org/)
186: - [OpenZFS on OS X ](https://openzfsonosx.org/) [repo](https://github.com/openzfsonosx)
187: - proprietary ZFS in Solaris (not relevant in open source)
188: - ZFS as released under the CDDL (common ancestor, now of historical interest)
189:
190: OpenZFS is a coordinating project to align open ZFS codebases. There
191: is a notion of a shared core codebase and OS-specific adaptation code.
192:
193: - [zfsonlinux relationship to OpenZFS](https://github.com/openzfs/zfs/wiki/OpenZFS-Patches)
194: - FreeBSD more or less imports code from openzfs and pushes back fixes. \verify
195: - NetBSD has imported code from FreeBSD
196: - The status of ZFS on macOS is unclear (2021-02)
CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb