Hi.
Thumpers come with Solaris pre-installed and already configured one pool.
It''s a collection of raid-z1 groups but some groups are smaller than
the others.
I''ll reconfigure it anyway but I''m just curious what
side-effects can there be with such a config?
Any performance hit? All space will be used?
# zpool status
pool: zpool1
state: ONLINE
status: The pool is formatted using an older on-disk format. The pool can
still be used, but some features are unavailable.
action: Upgrade the pool using ''zpool upgrade''. Once this is
done, the
pool will no longer be accessible on older software versions.
scrub: scrub stopped with 0 errors on Thu Nov 2 09:29:14 2006
config:
NAME STATE READ WRITE CKSUM
zpool1 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t0d0 ONLINE 0 0 0
c1t0d0 ONLINE 0 0 0
c4t0d0 ONLINE 0 0 0
c6t0d0 ONLINE 0 0 0
c7t0d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t1d0 ONLINE 0 0 0
c1t1d0 ONLINE 0 0 0
c4t1d0 ONLINE 0 0 0
c5t1d0 ONLINE 0 0 0
c6t1d0 ONLINE 0 0 0
c7t1d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t2d0 ONLINE 0 0 0
c1t2d0 ONLINE 0 0 0
c4t2d0 ONLINE 0 0 0
c5t2d0 ONLINE 0 0 0
c6t2d0 ONLINE 0 0 0
c7t2d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t3d0 ONLINE 0 0 0
c1t3d0 ONLINE 0 0 0
c4t3d0 ONLINE 0 0 0
c5t3d0 ONLINE 0 0 0
c6t3d0 ONLINE 0 0 0
c7t3d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t4d0 ONLINE 0 0 0
c1t4d0 ONLINE 0 0 0
c4t4d0 ONLINE 0 0 0
c6t4d0 ONLINE 0 0 0
c7t4d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t5d0 ONLINE 0 0 0
c1t5d0 ONLINE 0 0 0
c4t5d0 ONLINE 0 0 0
c5t5d0 ONLINE 0 0 0
c6t5d0 ONLINE 0 0 0
c7t5d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t6d0 ONLINE 0 0 0
c1t6d0 ONLINE 0 0 0
c4t6d0 ONLINE 0 0 0
c5t6d0 ONLINE 0 0 0
c6t6d0 ONLINE 0 0 0
c7t6d0 ONLINE 0 0 0
raidz1 ONLINE 0 0 0
c0t7d0 ONLINE 0 0 0
c1t7d0 ONLINE 0 0 0
c4t7d0 ONLINE 0 0 0
c5t7d0 ONLINE 0 0 0
c6t7d0 ONLINE 0 0 0
c7t7d0 ONLINE 0 0 0
errors: No known data errors
bash-3.00#
This message posted from opensolaris.org
Robert Milkowski wrote:> Thumpers come with Solaris pre-installed and already configured one pool. > It''s a collection of raid-z1 groups but some groups are smaller than the others. > I''ll reconfigure it anyway but I''m just curious what side-effects can there be with such a config? > Any performance hit? All space will be used?Here is some of our thoughts which led to the decision to pre-configure what you see on thumper. 1. Unallocated disk space doesn''t exist. You will notice that all other vendors who pre-install OSes on their systems ship with all of the space allocated. For example, buy a Mac, do a df and all of the space can be seen as allocated. Traditionally, Sun has delivered pre-installed systems with unallocated space and if you wanted to use it, you had to search for it, use format, etc. 2. It takes only one command to destroy the zpool. This fits with #1. I''m a strong advocate for #1. It is easier to destroy the pre-installed zpool and recreate it than to search for space using format (or whatever). This is especially true for a large JBOD like thumper. 3. The data should be protected by default. 4. The available space should be maximized, given constraint #3. Big-wide raidz1 sets are not a good idea, hence the multiple sets. Incidentally, while it doesn''t really make much difference, wrt data availability, which controller you use, we knew that people would freak out if we didn''t spread the vdevs across the controllers. 5. We''re not using a hot spare. This is a concession for #1. But if you truly value data availability and retention, use a hot spare. So, what you get, out of the box, is a large pool ready to be used. The pool has respectable performance and data protection. But, if you need to configure for more performance or data protection, then it is easy to do. P.S. did you upgrade the OS? I''d consider the need for ''zpool upgrade'' to be a bug. -- richard> # zpool status > pool: zpool1 > state: ONLINE > status: The pool is formatted using an older on-disk format. The pool can > still be used, but some features are unavailable. > action: Upgrade the pool using ''zpool upgrade''. Once this is done, the > pool will no longer be accessible on older software versions. > scrub: scrub stopped with 0 errors on Thu Nov 2 09:29:14 2006 > config: > > NAME STATE READ WRITE CKSUM > zpool1 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t0d0 ONLINE 0 0 0 > c1t0d0 ONLINE 0 0 0 > c4t0d0 ONLINE 0 0 0 > c6t0d0 ONLINE 0 0 0 > c7t0d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t1d0 ONLINE 0 0 0 > c1t1d0 ONLINE 0 0 0 > c4t1d0 ONLINE 0 0 0 > c5t1d0 ONLINE 0 0 0 > c6t1d0 ONLINE 0 0 0 > c7t1d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t2d0 ONLINE 0 0 0 > c1t2d0 ONLINE 0 0 0 > c4t2d0 ONLINE 0 0 0 > c5t2d0 ONLINE 0 0 0 > c6t2d0 ONLINE 0 0 0 > c7t2d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t3d0 ONLINE 0 0 0 > c1t3d0 ONLINE 0 0 0 > c4t3d0 ONLINE 0 0 0 > c5t3d0 ONLINE 0 0 0 > c6t3d0 ONLINE 0 0 0 > c7t3d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t4d0 ONLINE 0 0 0 > c1t4d0 ONLINE 0 0 0 > c4t4d0 ONLINE 0 0 0 > c6t4d0 ONLINE 0 0 0 > c7t4d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t5d0 ONLINE 0 0 0 > c1t5d0 ONLINE 0 0 0 > c4t5d0 ONLINE 0 0 0 > c5t5d0 ONLINE 0 0 0 > c6t5d0 ONLINE 0 0 0 > c7t5d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t6d0 ONLINE 0 0 0 > c1t6d0 ONLINE 0 0 0 > c4t6d0 ONLINE 0 0 0 > c5t6d0 ONLINE 0 0 0 > c6t6d0 ONLINE 0 0 0 > c7t6d0 ONLINE 0 0 0 > raidz1 ONLINE 0 0 0 > c0t7d0 ONLINE 0 0 0 > c1t7d0 ONLINE 0 0 0 > c4t7d0 ONLINE 0 0 0 > c5t7d0 ONLINE 0 0 0 > c6t7d0 ONLINE 0 0 0 > c7t7d0 ONLINE 0 0 0 > > errors: No known data errors > bash-3.00# > > > This message posted from opensolaris.org > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Hello Richard, Thursday, November 2, 2006, 7:08:17 PM, you wrote: REP> Robert Milkowski wrote:>> Thumpers come with Solaris pre-installed and already configured one pool. >> It''s a collection of raid-z1 groups but some groups are smaller than the others. >> I''ll reconfigure it anyway but I''m just curious what side-effects can there be with such a config? >> Any performance hit? All space will be used?REP> Here is some of our thoughts which led to the decision to pre-configure REP> what you see on thumper. REP> 1. Unallocated disk space doesn''t exist. You will notice that all other REP> vendors who pre-install OSes on their systems ship with all of the space REP> allocated. For example, buy a Mac, do a df and all of the space can be REP> seen as allocated. Traditionally, Sun has delivered pre-installed systems REP> with unallocated space and if you wanted to use it, you had to search for REP> it, use format, etc. REP> 2. It takes only one command to destroy the zpool. This fits with #1. I''m REP> a strong advocate for #1. It is easier to destroy the pre-installed zpool REP> and recreate it than to search for space using format (or whatever). This REP> is especially true for a large JBOD like thumper. REP> 3. The data should be protected by default. REP> 4. The available space should be maximized, given constraint #3. Big-wide REP> raidz1 sets are not a good idea, hence the multiple sets. Incidentally, REP> while it doesn''t really make much difference, wrt data availability, which REP> controller you use, we knew that people would freak out if we didn''t spread REP> the vdevs across the controllers. REP> 5. We''re not using a hot spare. This is a concession for #1. But if you REP> truly value data availability and retention, use a hot spare. REP> So, what you get, out of the box, is a large pool ready to be used. The pool REP> has respectable performance and data protection. But, if you need to configure REP> for more performance or data protection, then it is easy to do. REP> P.S. did you upgrade the OS? I''d consider the need for ''zpool upgrade'' to be REP> a bug. on one thumper I reinstalled OS to S10U3 beta and imported default pool. On another I put snv_49 and imported pool. Then I destroyed pools and I''m experimenting with different configurations. I almost completely agree with your points 1-5, except that I think that having at least one hot spare by default would be better than having none at all - especially with SATA drives. I''ll probably go with a config similar to default pool but with 2 hot spares. ps. and you (Sun) haven''t configured hot spare as Thumper so far comes with U2 and there are no hot spares in U2 (and for strange reason thumper also comes with U1 - at least it''s in GRUB menu I haven''t tried to boot it - will try with another Thumper). -- Best regards, Robert mailto:rmilkowski at task.gda.pl http://milek.blogspot.com
Robert Milkowski wrote:> REP> P.S. did you upgrade the OS? I''d consider the need for ''zpool upgrade'' to be > REP> a bug. > > on one thumper I reinstalled OS to S10U3 beta and imported default > pool. On another I put snv_49 and imported pool. Then I destroyed > pools and I''m experimenting with different configurations.ah, ok.> I almost completely agree with your points 1-5, except that I think > that having at least one hot spare by default would be better than > having none at all - especially with SATA drives.Yes, I pushed for it, but didn''t win.> I''ll probably go with a config similar to default pool but with 2 hot > spares.I heartily recommend hot spares.> ps. and you (Sun) haven''t configured hot spare as Thumper so far comes > with U2 and there are no hot spares in U2 (and for strange reason > thumper also comes with U1 - at least it''s in GRUB > menu I haven''t tried to boot it - will try with another Thumper). >That isn''t really surprising. New hardware tends to drive features into the OS. When the two are releasing at different rates, then you get into some bootstrapping issues. Once we release U3 (aka Solaris 10 11/06) then we can cutover to that in the factory (can''t release GA hardware with non-GA software) and soon the both with be in proper sync. -- richard
Richard Elling - PAE wrote:> Robert Milkowski wrote: > >> I almost completely agree with your points 1-5, except that I think >> that having at least one hot spare by default would be better than >> having none at all - especially with SATA drives. > > Yes, I pushed for it, but didn''t win.In a perfect world one could simply pull one of the raidz1 groups out of the pool and allocate hot spares out of it. That way you''re one or two commands away from the "most space config" to "lots of redundancy" config. Not that recreating the pool is a lot of work but I can see some folks just using the box for awhile and then thinking, "Hey....no hot spares?"