A slightly different tack now... what filesystems is it a good (or bad) idea to put on ZFS? root - NO (not yet anyway) home - YES (although the huge number of mounts still scares me a bit) /usr - possible? /var - possible? swap - no? Is there any advantage in having multiple zpools over just having one and allocating all filesystems out of it? Obviously if you wanted (for example) /export/home to be raidz and /usr to be mirror you would have to, but are there other considerations beyond that? I''m thinking that zfs frees me up from getting the sizing ''right'' at install time i.e. big enough that I don''t have to resize later, which inevitably means at least one filesystem being far bigger than it needs to be. Steve.
Bennett, Steve wrote:> A slightly different tack now... > > what filesystems is it a good (or bad) idea to put on ZFS? > root - NO (not yet anyway) > home - YES (although the huge number of mounts still scares me a bit) > /usr - possible?not yet - the system wouldn''t be patchable or upgradeable.> /var - possible?not yet, same reasons.> swap - no?for now, maybe. Offhand I''d say you should have some swap on a slice for now, although you can have additional swap space on zvols. When zfs root is supported, the swap space set up by install will be a zvol in the root pool.> > Is there any advantage in having multiple zpools over just having one > and allocating all filesystems out of it? > Obviously if you wanted (for example) /export/home to be raidz and /usr > to be mirror you would have to, but are there other considerations > beyond that?It is likely that "best practice" will be to separate the root pool (that is, the pool where dataset are allocated for the purpose of serving as root file systems) from pool(s) used for data. There are several reasons for this: * there will be some limitations on root pools that you would not want to place on data pools. For example, no concatenation or RAID-Z (mirroring will be supported, however). * data pools can be architecture-neutral. It can make sense to move a data pool between sparc and intel. Root pools are pretty much tied to a particular architecture. * In general, we think it''s a good idea to separate the "personality" of a system from its data. Then you can change one without having to change the other. But separate pools for what we now sometimes allocate as separate file systems (root, /usr and /var, for example) makes no sense at all. It probably won''t even be a supported configuration. Separate *datasets* for those directories will be possible, but only in the same pool.> > I''m thinking that zfs frees me up from getting the sizing ''right'' at > install time i.e. big enough that I don''t have to resize later, which > inevitably means at least one filesystem being far bigger than it needs > to be. >Yes, that''s one of the advantages of zfs we hope to take advantage of in install procedures when zfs is supported as a root file system. Lori
Bennett, Steve wrote:> A slightly different tack now... > > what filesystems is it a good (or bad) idea to put on ZFS? > root - NO (not yet anyway) > home - YES (although the huge number of mounts still scares me a bit)Why ? I''ve seen Solaris systems with upwards of 10,000 mounts in /home via the automounter alone.> /usr - possible? > /var - possible? > swap - no?You can swap on a zvol today but note that you can''t yet use a zvol as a dump device.> Is there any advantage in having multiple zpools over just having one > and allocating all filesystems out of it? > Obviously if you wanted (for example) /export/home to be raidz and /usr > to be mirror you would have to, but are there other considerations > beyond that?Having a separate pool for the operating system and one for your data might be a good idea. This allows you to move the data else where later much easier. -- Darren J Moffat
> It is likely that "best practice" will be to separate > the root pool (that is, the pool where dataset are > allocated)On a system with plenty of disks it is a good idea. I started doing this on my laptop, and later decided to combine root and data into one pool. The filesystem boundary gave me sufficent separation. Having separate pools made me have 2 partitions with fixed boundries, which limited ZFS''s flexibility. Doug This message posted from opensolaris.org
Doug Scott wrote:>>It is likely that "best practice" will be to separate >>the root pool (that is, the pool where dataset are >>allocated) > > On a system with plenty of disks it is a good idea. I started > doing this on my laptop, and later decided to combine root and > data into one pool. The filesystem boundary gave me sufficent > separation. Having separate pools made me have 2 partitions > with fixed boundries, which limited ZFS''s flexibility.For example, ISTR that pools can be exported, but individual filesystems within a pool cannot be exported separately from the pool in which they are located. -- -------------------------------------------------------------------------- Jeff VICTOR Sun Microsystems jeff.victor @ sun.com OS Ambassador Sr. Technical Specialist Solaris 10 Zones FAQ: http://www.opensolaris.org/os/community/zones/faq --------------------------------------------------------------------------
> Doug Scott wrote: > >>It is likely that "best practice" will be to > separate > >>the root pool (that is, the pool where dataset are > >>allocated) > > > > On a system with plenty of disks it is a good idea. > I started > > doing this on my laptop, and later decided to > combine root and > > data into one pool. The filesystem boundary gave me > sufficent > > separation. Having separate pools made me have 2 > partitions > > with fixed boundries, which limited ZFS''s > flexibility. > > For example, ISTR that pools can be exported, but > individual filesystems within a > pool cannot be exported separately from the pool in > which they are located.That is correct, but on my laptop I have little need to export the pool. Since diskspace is a limited resource I prefer that I did not have fixed partitions, and let all filesystems be able to access the remaining space. I hit the a partition boundy the first time I cloned more root zfs filesystem, thus I changed to a single pool. I can still use ''zfs send'' to copy any of the filesystems elsewhere. Doug This message posted from opensolaris.org