zfs discuss - Oct 2011 - weird bug with Seagate 3TB USB3 drive

Banging my head against a Seagate 3TB USB3 drive.
Its marketing name is:
Seagate Expansion 3 TB USB 3.0 Desktop External Hard Drive STAY3000102
format(1M) shows it identify itself as:
Seagate-External-SG11-2.73TB

Under both Solaris 10 and Solaris 11x, I receive the evil message:
| I/O request is not aligned with 4096 disk sector size.
| It is handled through Read Modify Write but the performance is very low.

However, that''s not my big issue as I will use the zpool-12 hack.

My big issue is that once I zpool(1M) export the pool from
my W2100z running S10 or my Ultra 40 running S11x, I can''t 
import it.

I thought weird USB connectivity issue, but I can run
"format -> analyze -> read" merrily.

Anyone seen this bug?

John
groenveld at acm.org

Hi John,

What is the error when you attempt to import this pool?

Thanks,

Cindy

On 10/11/11 18:17, John D Groenveld wrote:
> Banging my head against a Seagate 3TB USB3 drive.
> Its marketing name is:
> Seagate Expansion 3 TB USB 3.0 Desktop External Hard Drive STAY3000102
> format(1M) shows it identify itself as:
> Seagate-External-SG11-2.73TB
> 
> Under both Solaris 10 and Solaris 11x, I receive the evil message:
> | I/O request is not aligned with 4096 disk sector size.
> | It is handled through Read Modify Write but the performance is very low.
> 
> However, that''s not my big issue as I will use the zpool-12 hack.
> 
> My big issue is that once I zpool(1M) export the pool from
> my W2100z running S10 or my Ultra 40 running S11x, I can''t 
> import it.
> 
> I thought weird USB connectivity issue, but I can run
> "format -> analyze -> read" merrily.
> 
> Anyone seen this bug?
> 
> John
> groenveld at acm.org
> 
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <4E95CB2A.30105 at oracle.com>, Cindy Swearingen
writes:
>What is the error when you attempt to import this pool?
"cannot import ''foo'': no such pool available"
John
groenveld at acm.org

# format -e
Searching for disks...done


AVAILABLE DISK SELECTIONS:
       0. c1t0d0 <Seagate-External-SG11 cyl 45597 alt 2 hd 255 sec 63>
          /pci at 0,0/pci108e,6676 at 2,1/hub at 7/storage at 2/disk at 0,0
       1. c8t0d0 <ATA    -HITACHI HDS7225-A9CA cyl 30397 alt 2 hd 255 sec
63>
          /pci at 0,0/pci108e,6676 at 5/disk at 0,0
       2. c8t1d0 <ATA    -HITACHI HDS7225-A7BA cyl 30397 alt 2 hd 255 sec
63>
          /pci at 0,0/pci108e,6676 at 5/disk at 1,0
Specify disk (enter its number): ^C
# zpool create foo c1t0d0
# zfs create foo/bar
# zfs list -r foo
NAME      USED  AVAIL  REFER  MOUNTPOINT
foo       126K  2.68T    32K  /foo
foo/bar    31K  2.68T    31K  /foo/bar
# zpool export foo
# zfs list -r foo
cannot open ''foo'': dataset does not exist
# truss -t open zpool import foo
open("/var/ld/ld.config", O_RDONLY)             Err#2 ENOENT
open("/lib/libumem.so.1", O_RDONLY)             = 3
open("/lib/libc.so.1", O_RDONLY)                = 3
open("/lib/libzfs.so.1", O_RDONLY)              = 3
open("/usr/lib/fm//libtopo.so", O_RDONLY)       = 3
open("/lib/libxml2.so.2", O_RDONLY)             = 3
open("/lib/libpthread.so.1", O_RDONLY)          = 3
open("/lib/libz.so.1", O_RDONLY)                = 3
open("/lib/libm.so.2", O_RDONLY)                = 3
open("/lib/libsocket.so.1", O_RDONLY)           = 3
open("/lib/libnsl.so.1", O_RDONLY)              = 3
open("/usr/lib//libshare.so.1", O_RDONLY)       = 3
open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_SGS.mo",
O_RDONLY) Err#2 ENOENT
open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSLIB.mo",
O_RDONLY) Err#2 ENOENT
open("/usr/lib/locale/en_US.UTF-8/en_US.UTF-8.so.3", O_RDONLY) = 3
open("/usr/lib/locale/en_US.UTF-8/methods_unicode.so.3", O_RDONLY) = 3
open("/dev/zfs", O_RDWR)                        = 3
open("/etc/mnttab", O_RDONLY)                   = 4
open("/etc/dfs/sharetab", O_RDONLY)             = 5
open("/lib/libavl.so.1", O_RDONLY)              = 6
open("/lib/libnvpair.so.1", O_RDONLY)           = 6
open("/lib/libuutil.so.1", O_RDONLY)            = 6
open64("/dev/rdsk/", O_RDONLY)                  = 6
/3:     openat64(6, "c8t0d0s0", O_RDONLY)               = 9
/3:     open("/lib/libadm.so.1", O_RDONLY)              = 15
/9:     openat64(6, "c8t0d0s2", O_RDONLY)               = 13
/5:     openat64(6, "c8t1d0s0", O_RDONLY)               = 10
/7:     openat64(6, "c8t1d0s2", O_RDONLY)               = 14
/8:     openat64(6, "c1t0d0s0", O_RDONLY)               = 7
/4:     openat64(6, "c1t0d0s2", O_RDONLY)               Err#5 EIO
/8:     open("/lib/libefi.so.1", O_RDONLY)              = 15
/3:     openat64(6, "c1t0d0", O_RDONLY)                 = 9
/5:     openat64(6, "c1t0d0p0", O_RDONLY)               = 10
/9:     openat64(6, "c1t0d0p1", O_RDONLY)               = 13
/7:     openat64(6, "c1t0d0p2", O_RDONLY)               Err#5 EIO
/4:     openat64(6, "c1t0d0p3", O_RDONLY)               Err#5 EIO
/7:     openat64(6, "c1t0d0s8", O_RDONLY)               = 14
/2:     openat64(6, "c7t0d0s0", O_RDONLY)               = 8
/6:     openat64(6, "c7t0d0s2", O_RDONLY)               = 12
/1:         Received signal #20, SIGWINCH, in lwp_park() [default]
/3:     openat64(6, "c7t0d0p0", O_RDONLY)               = 9
/4:     openat64(6, "c7t0d0p1", O_RDONLY)               = 11
/5:     openat64(6, "c7t0d0p2", O_RDONLY)               = 10
/6:     openat64(6, "c8t0d0p0", O_RDONLY)               = 12
/6:     openat64(6, "c8t0d0p1", O_RDONLY)               = 12
/6:     openat64(6, "c8t0d0p2", O_RDONLY)               Err#5 EIO
/6:     openat64(6, "c8t0d0p3", O_RDONLY)               Err#5 EIO
/6:     openat64(6, "c8t0d0p4", O_RDONLY)               Err#5 EIO
/6:     openat64(6, "c8t1d0p0", O_RDONLY)               = 12
/8:     openat64(6, "c7t0d0p3", O_RDONLY)               = 7
/6:     openat64(6, "c8t1d0p1", O_RDONLY)               = 12
/6:     openat64(6, "c8t1d0p2", O_RDONLY)               Err#5 EIO
/6:     openat64(6, "c8t1d0p3", O_RDONLY)               Err#5 EIO
/6:     openat64(6, "c8t1d0p4", O_RDONLY)               Err#5 EIO
/9:     openat64(6, "c7t0d0p4", O_RDONLY)               = 13
/7:     openat64(6, "c7t0d0s1", O_RDONLY)               = 14
/1:    
open("/usr/share/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.cat",
O_RDONLY) Err#2 ENOENT
open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.mo",
O_RDONLY) Err#2 ENOENT
cannot import ''foo'': no such pool available

In the steps below, you''re missing a zpool import step.
I would like to see the error message when the zpool import
step fails.

Thanks,

Cindy

On 10/12/11 11:29, John D Groenveld wrote:
> In message <4E95CB2A.30105 at oracle.com>, Cindy Swearingen writes:
>> What is the error when you attempt to import this pool?
> 
> "cannot import ''foo'': no such pool available"
> John
> groenveld at acm.org
> 
> # format -e
> Searching for disks...done
> 
> 
> AVAILABLE DISK SELECTIONS:
>        0. c1t0d0 <Seagate-External-SG11 cyl 45597 alt 2 hd 255 sec
63>
>           /pci at 0,0/pci108e,6676 at 2,1/hub at 7/storage at 2/disk at 0,0
>        1. c8t0d0 <ATA    -HITACHI HDS7225-A9CA cyl 30397 alt 2 hd 255
sec 63>
>           /pci at 0,0/pci108e,6676 at 5/disk at 0,0
>        2. c8t1d0 <ATA    -HITACHI HDS7225-A7BA cyl 30397 alt 2 hd 255
sec 63>
>           /pci at 0,0/pci108e,6676 at 5/disk at 1,0
> Specify disk (enter its number): ^C
> # zpool create foo c1t0d0
> # zfs create foo/bar
> # zfs list -r foo
> NAME      USED  AVAIL  REFER  MOUNTPOINT
> foo       126K  2.68T    32K  /foo
> foo/bar    31K  2.68T    31K  /foo/bar
> # zpool export foo
> # zfs list -r foo
> cannot open ''foo'': dataset does not exist
> # truss -t open zpool import foo
> open("/var/ld/ld.config", O_RDONLY)             Err#2 ENOENT
> open("/lib/libumem.so.1", O_RDONLY)             = 3
> open("/lib/libc.so.1", O_RDONLY)                = 3
> open("/lib/libzfs.so.1", O_RDONLY)              = 3
> open("/usr/lib/fm//libtopo.so", O_RDONLY)       = 3
> open("/lib/libxml2.so.2", O_RDONLY)             = 3
> open("/lib/libpthread.so.1", O_RDONLY)          = 3
> open("/lib/libz.so.1", O_RDONLY)                = 3
> open("/lib/libm.so.2", O_RDONLY)                = 3
> open("/lib/libsocket.so.1", O_RDONLY)           = 3
> open("/lib/libnsl.so.1", O_RDONLY)              = 3
> open("/usr/lib//libshare.so.1", O_RDONLY)       = 3
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_SGS.mo",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSLIB.mo",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/en_US.UTF-8.so.3", O_RDONLY) =
3
> open("/usr/lib/locale/en_US.UTF-8/methods_unicode.so.3",
O_RDONLY) = 3
> open("/dev/zfs", O_RDWR)                        = 3
> open("/etc/mnttab", O_RDONLY)                   = 4
> open("/etc/dfs/sharetab", O_RDONLY)             = 5
> open("/lib/libavl.so.1", O_RDONLY)              = 6
> open("/lib/libnvpair.so.1", O_RDONLY)           = 6
> open("/lib/libuutil.so.1", O_RDONLY)            = 6
> open64("/dev/rdsk/", O_RDONLY)                  = 6
> /3:     openat64(6, "c8t0d0s0", O_RDONLY)               = 9
> /3:     open("/lib/libadm.so.1", O_RDONLY)              = 15
> /9:     openat64(6, "c8t0d0s2", O_RDONLY)               = 13
> /5:     openat64(6, "c8t1d0s0", O_RDONLY)               = 10
> /7:     openat64(6, "c8t1d0s2", O_RDONLY)               = 14
> /8:     openat64(6, "c1t0d0s0", O_RDONLY)               = 7
> /4:     openat64(6, "c1t0d0s2", O_RDONLY)               Err#5 EIO
> /8:     open("/lib/libefi.so.1", O_RDONLY)              = 15
> /3:     openat64(6, "c1t0d0", O_RDONLY)                 = 9
> /5:     openat64(6, "c1t0d0p0", O_RDONLY)               = 10
> /9:     openat64(6, "c1t0d0p1", O_RDONLY)               = 13
> /7:     openat64(6, "c1t0d0p2", O_RDONLY)               Err#5 EIO
> /4:     openat64(6, "c1t0d0p3", O_RDONLY)               Err#5 EIO
> /7:     openat64(6, "c1t0d0s8", O_RDONLY)               = 14
> /2:     openat64(6, "c7t0d0s0", O_RDONLY)               = 8
> /6:     openat64(6, "c7t0d0s2", O_RDONLY)               = 12
> /1:         Received signal #20, SIGWINCH, in lwp_park() [default]
> /3:     openat64(6, "c7t0d0p0", O_RDONLY)               = 9
> /4:     openat64(6, "c7t0d0p1", O_RDONLY)               = 11
> /5:     openat64(6, "c7t0d0p2", O_RDONLY)               = 10
> /6:     openat64(6, "c8t0d0p0", O_RDONLY)               = 12
> /6:     openat64(6, "c8t0d0p1", O_RDONLY)               = 12
> /6:     openat64(6, "c8t0d0p2", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t0d0p3", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t0d0p4", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p0", O_RDONLY)               = 12
> /8:     openat64(6, "c7t0d0p3", O_RDONLY)               = 7
> /6:     openat64(6, "c8t1d0p1", O_RDONLY)               = 12
> /6:     openat64(6, "c8t1d0p2", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p3", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p4", O_RDONLY)               Err#5 EIO
> /9:     openat64(6, "c7t0d0p4", O_RDONLY)               = 13
> /7:     openat64(6, "c7t0d0s1", O_RDONLY)               = 14
> /1:    
open("/usr/share/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.cat",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.mo",
O_RDONLY) Err#2 ENOENT
> cannot import ''foo'': no such pool available
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <4E95E10F.9070108 at oracle.com>, Cindy Swearingen
writes:
>In the steps below, you''re missing a zpool import step.
>I would like to see the error message when the zpool import
>step fails.
"zpool import" returns nothing.
The truss shows it poking around c1t0d0 fdisk partitions and
Solaris slices presumably hunting for pools.

John
groenveld at acm.org

> From: zfs-discuss-bounces at opensolaris.org [mailto:zfs-discuss-
> bounces at opensolaris.org] On Behalf Of Cindy Swearingen
> 
> In the steps below, you''re missing a zpool import step.
> I would like to see the error message when the zpool import
> step fails.
I see him doing this...

> > # truss -t open zpool import foo
The following lines are informative, sort of.

> > /8:     openat64(6, "c1t0d0s0", O_RDONLY)               = 7
> > /4:     openat64(6, "c1t0d0s2", O_RDONLY)              
Err#5 EIO
And the output result is:

> > cannot import ''foo'': no such pool available

>> From: zfs-discuss-bounces at opensolaris.org [mailto:zfs-discuss-
>> bounces at opensolaris.org] On Behalf Of Cindy Swearingen
>> 
>> In the steps below, you''re missing a zpool import step.
>> I would like to see the error message when the zpool import
>> step fails.
>
>I see him doing this...
>
>
>> > # truss -t open zpool import foo
>
>The following lines are informative, sort of.
>
>
>> > /8:     openat64(6, "c1t0d0s0", O_RDONLY)              
= 7
>> > /4:     openat64(6, "c1t0d0s2", O_RDONLY)              
Err#5 EIO
>
>And the output result is:
>
>
>> > cannot import ''foo'': no such pool available
>
What is the partition table?

Casper

> From: Casper.Dik at oracle.com [mailto:Casper.Dik at oracle.com]
> 
> What is the partition table?
He also said this...

> -----Original Message-----
> From: zfs-discuss-bounces at opensolaris.org [mailto:zfs-discuss-
> bounces at opensolaris.org] On Behalf Of John D Groenveld
>
> # zpool create foo c1t0d0
Which, to me, suggests no partition table.

>> From: Casper.Dik at oracle.com [mailto:Casper.Dik at oracle.com]
>> 
>> What is the partition table?
>
>He also said this...
>
>
>> -----Original Message-----
>> From: zfs-discuss-bounces at opensolaris.org [mailto:zfs-discuss-
>> bounces at opensolaris.org] On Behalf Of John D Groenveld
>>
>> # zpool create foo c1t0d0
>
>Which, to me, suggests no partition table.
An EFI partition table (there needs to be some form of label so there
is always a partition table).

Casper

John,

Any USB-related messages in /var/adm/messages for this device?

Thanks,

Cindy

On 10/12/11 11:29, John D Groenveld wrote:
> In message <4E95CB2A.30105 at oracle.com>, Cindy Swearingen writes:
>> What is the error when you attempt to import this pool?
> 
> "cannot import ''foo'': no such pool available"
> John
> groenveld at acm.org
> 
> # format -e
> Searching for disks...done
> 
> 
> AVAILABLE DISK SELECTIONS:
>        0. c1t0d0 <Seagate-External-SG11 cyl 45597 alt 2 hd 255 sec
63>
>           /pci at 0,0/pci108e,6676 at 2,1/hub at 7/storage at 2/disk at 0,0
>        1. c8t0d0 <ATA    -HITACHI HDS7225-A9CA cyl 30397 alt 2 hd 255
sec 63>
>           /pci at 0,0/pci108e,6676 at 5/disk at 0,0
>        2. c8t1d0 <ATA    -HITACHI HDS7225-A7BA cyl 30397 alt 2 hd 255
sec 63>
>           /pci at 0,0/pci108e,6676 at 5/disk at 1,0
> Specify disk (enter its number): ^C
> # zpool create foo c1t0d0
> # zfs create foo/bar
> # zfs list -r foo
> NAME      USED  AVAIL  REFER  MOUNTPOINT
> foo       126K  2.68T    32K  /foo
> foo/bar    31K  2.68T    31K  /foo/bar
> # zpool export foo
> # zfs list -r foo
> cannot open ''foo'': dataset does not exist
> # truss -t open zpool import foo
> open("/var/ld/ld.config", O_RDONLY)             Err#2 ENOENT
> open("/lib/libumem.so.1", O_RDONLY)             = 3
> open("/lib/libc.so.1", O_RDONLY)                = 3
> open("/lib/libzfs.so.1", O_RDONLY)              = 3
> open("/usr/lib/fm//libtopo.so", O_RDONLY)       = 3
> open("/lib/libxml2.so.2", O_RDONLY)             = 3
> open("/lib/libpthread.so.1", O_RDONLY)          = 3
> open("/lib/libz.so.1", O_RDONLY)                = 3
> open("/lib/libm.so.2", O_RDONLY)                = 3
> open("/lib/libsocket.so.1", O_RDONLY)           = 3
> open("/lib/libnsl.so.1", O_RDONLY)              = 3
> open("/usr/lib//libshare.so.1", O_RDONLY)       = 3
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_SGS.mo",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSLIB.mo",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/en_US.UTF-8.so.3", O_RDONLY) =
3
> open("/usr/lib/locale/en_US.UTF-8/methods_unicode.so.3",
O_RDONLY) = 3
> open("/dev/zfs", O_RDWR)                        = 3
> open("/etc/mnttab", O_RDONLY)                   = 4
> open("/etc/dfs/sharetab", O_RDONLY)             = 5
> open("/lib/libavl.so.1", O_RDONLY)              = 6
> open("/lib/libnvpair.so.1", O_RDONLY)           = 6
> open("/lib/libuutil.so.1", O_RDONLY)            = 6
> open64("/dev/rdsk/", O_RDONLY)                  = 6
> /3:     openat64(6, "c8t0d0s0", O_RDONLY)               = 9
> /3:     open("/lib/libadm.so.1", O_RDONLY)              = 15
> /9:     openat64(6, "c8t0d0s2", O_RDONLY)               = 13
> /5:     openat64(6, "c8t1d0s0", O_RDONLY)               = 10
> /7:     openat64(6, "c8t1d0s2", O_RDONLY)               = 14
> /8:     openat64(6, "c1t0d0s0", O_RDONLY)               = 7
> /4:     openat64(6, "c1t0d0s2", O_RDONLY)               Err#5 EIO
> /8:     open("/lib/libefi.so.1", O_RDONLY)              = 15
> /3:     openat64(6, "c1t0d0", O_RDONLY)                 = 9
> /5:     openat64(6, "c1t0d0p0", O_RDONLY)               = 10
> /9:     openat64(6, "c1t0d0p1", O_RDONLY)               = 13
> /7:     openat64(6, "c1t0d0p2", O_RDONLY)               Err#5 EIO
> /4:     openat64(6, "c1t0d0p3", O_RDONLY)               Err#5 EIO
> /7:     openat64(6, "c1t0d0s8", O_RDONLY)               = 14
> /2:     openat64(6, "c7t0d0s0", O_RDONLY)               = 8
> /6:     openat64(6, "c7t0d0s2", O_RDONLY)               = 12
> /1:         Received signal #20, SIGWINCH, in lwp_park() [default]
> /3:     openat64(6, "c7t0d0p0", O_RDONLY)               = 9
> /4:     openat64(6, "c7t0d0p1", O_RDONLY)               = 11
> /5:     openat64(6, "c7t0d0p2", O_RDONLY)               = 10
> /6:     openat64(6, "c8t0d0p0", O_RDONLY)               = 12
> /6:     openat64(6, "c8t0d0p1", O_RDONLY)               = 12
> /6:     openat64(6, "c8t0d0p2", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t0d0p3", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t0d0p4", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p0", O_RDONLY)               = 12
> /8:     openat64(6, "c7t0d0p3", O_RDONLY)               = 7
> /6:     openat64(6, "c8t1d0p1", O_RDONLY)               = 12
> /6:     openat64(6, "c8t1d0p2", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p3", O_RDONLY)               Err#5 EIO
> /6:     openat64(6, "c8t1d0p4", O_RDONLY)               Err#5 EIO
> /9:     openat64(6, "c7t0d0p4", O_RDONLY)               = 13
> /7:     openat64(6, "c7t0d0s1", O_RDONLY)               = 14
> /1:    
open("/usr/share/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.cat",
O_RDONLY) Err#2 ENOENT
> open("/usr/lib/locale/en_US.UTF-8/LC_MESSAGES/SUNW_OST_OSCMD.mo",
O_RDONLY) Err#2 ENOENT
> cannot import ''foo'': no such pool available
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <201110131150.p9DBo8Yk011167 at acsinet22.oracle.com>,
Casper.Dik at oracl
e.com writes:
>What is the partition table?
I thought about that so I reproduced with the legacy SMI label
and a Solaris fdisk partition with ZFS on slice 0.
Same result as EFI; once I export the pool I cannot import it.

John
groenveld at acm.org

In message <4E970387.3040800 at oracle.com>, Cindy Swearingen
writes:
>Any USB-related messages in /var/adm/messages for this device?
Negative.
cfgadm(1M) shows the drive and format->fdisk->analyze->read
runs merrily.

John
groenveld at acm.org

As a sanity check, I connected the drive to a Windows 7 installation.
I was able to partition, create an NTFS volume on it, eject and
remount it.

I also tried creating the zpool on my Solaris 10 system, exporting
and trying to import the pool on my Solaris 11X system and again
no love.

I''m baffled why zpool import is unable to find the pool on the
drive, but the drive is definitely functional.

John
groenveld at acm.org

On Oct 14, 2011, at 7:02 PM, John D Groenveld wrote:
> As a sanity check, I connected the drive to a Windows 7 installation.
> I was able to partition, create an NTFS volume on it, eject and
> remount it.
> 
> I also tried creating the zpool on my Solaris 10 system, exporting
> and trying to import the pool on my Solaris 11X system and again
> no love.
> 
> I''m baffled why zpool import is unable to find the pool on the
> drive, but the drive is definitely functional.
One of the best troubleshooting steps for a pool that won''t import is
to look at the labels on the disk.
	zdb -l /dev/dsk/c1t0d0s0

The output should be the nvlist for each of 4 labels on the device.
zpool import looks at those labels to determine if there is a pool available
for import. If the labels cannot be seen, then you need to solve that problem
before you can import the pool.
 -- richard

-- 

ZFS and performance consulting
http://www.RichardElling.com
VMworld Copenhagen, October 17-20
OpenStorage Summit, San Jose, CA, October 24-27
LISA ''11, Boston, MA, December 4-9

On Fri, 14 Oct 2011, John D Groenveld wrote:
>
> I''m baffled why zpool import is unable to find the pool on the
> drive, but the drive is definitely functional.
What type of controller is this drive attached to?  I have heard that 
some popular LSI controllers just don''t work for large drives.

Bob
-- 
Bob Friesenhahn
bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/

In message <201110150202.p9F22W2n000693 at elvis.arl.psu.edu>, John D
Groenveld writes:
>I''m baffled why zpool import is unable to find the pool on the
>drive, but the drive is definitely functional.
Per Richard Elling, it looks like ZFS is unable to find
the requisite labels for importing.

John
groenveld at acm.org

# prtvtoc /dev/rdsk/c1t0d0s2
* /dev/rdsk/c1t0d0s2 partition map
*
* Dimensions:
*    4096 bytes/sector
*      63 sectors/track
*     255 tracks/cylinder
*   16065 sectors/cylinder
*   45599 cylinders
*   45597 accessible cylinders
*
* Flags:
*   1: unmountable
*  10: read-only
*
* Unallocated space:
*       First     Sector    Last
*       Sector     Count    Sector
*           0     16065     16064
*
*                          First     Sector    Last
* Partition  Tag  Flags    Sector     Count    Sector  Mount Directory
       0      2    00      16065 732483675 732499739
       2      5    01          0 732515805 732515804
       8      1    01          0     16065     16064
# zpool create -f foobar c1t0d0s0
# zpool status foobar
  pool: foobar
 state: ONLINE
 scan: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        foobar      ONLINE       0     0     0
          c1t0d0s0  ONLINE       0     0     0

errors: No known data errors
# zdb -l /dev/dsk/c1t0d0s0
--------------------------------------------
LABEL 0
--------------------------------------------
failed to unpack label 0
--------------------------------------------
LABEL 1
--------------------------------------------
failed to unpack label 1
--------------------------------------------
LABEL 2
--------------------------------------------
failed to unpack label 2
--------------------------------------------
LABEL 3
--------------------------------------------
failed to unpack label 3

Hi John,

I''m going to file a CR to get this issue reviewed by the USB team
first, but if you could humor me with another test:

Can you run newfs to create a UFS file system on this device
and mount it?

Thanks,

Cindy

On 10/18/11 08:18, John D Groenveld wrote:
> In message <201110150202.p9F22W2n000693 at elvis.arl.psu.edu>, John D
Groenveld writes:
>> I''m baffled why zpool import is unable to find the pool on the
>> drive, but the drive is definitely functional.
> 
> Per Richard Elling, it looks like ZFS is unable to find
> the requisite labels for importing.
> 
> John
> groenveld at acm.org
> 
> # prtvtoc /dev/rdsk/c1t0d0s2
> * /dev/rdsk/c1t0d0s2 partition map
> *
> * Dimensions:
> *    4096 bytes/sector
> *      63 sectors/track
> *     255 tracks/cylinder
> *   16065 sectors/cylinder
> *   45599 cylinders
> *   45597 accessible cylinders
> *
> * Flags:
> *   1: unmountable
> *  10: read-only
> *
> * Unallocated space:
> *       First     Sector    Last
> *       Sector     Count    Sector
> *           0     16065     16064
> *
> *                          First     Sector    Last
> * Partition  Tag  Flags    Sector     Count    Sector  Mount Directory
>        0      2    00      16065 732483675 732499739
>        2      5    01          0 732515805 732515804
>        8      1    01          0     16065     16064
> # zpool create -f foobar c1t0d0s0
> # zpool status foobar
>   pool: foobar
>  state: ONLINE
>  scan: none requested
> config:
> 
>         NAME        STATE     READ WRITE CKSUM
>         foobar      ONLINE       0     0     0
>           c1t0d0s0  ONLINE       0     0     0
> 
> errors: No known data errors
> # zdb -l /dev/dsk/c1t0d0s0
> --------------------------------------------
> LABEL 0
> --------------------------------------------
> failed to unpack label 0
> --------------------------------------------
> LABEL 1
> --------------------------------------------
> failed to unpack label 1
> --------------------------------------------
> LABEL 2
> --------------------------------------------
> failed to unpack label 2
> --------------------------------------------
> LABEL 3
> --------------------------------------------
> failed to unpack label 3
> 
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <4E9D98B1.8040108 at oracle.com>, Cindy Swearingen
writes:
>I''m going to file a CR to get this issue reviewed by the USB team
>first, but if you could humor me with another test:
>
>Can you run newfs to create a UFS file system on this device
>and mount it?
# uname -srvp
SunOS 5.11 151.0.1.12 i386
# zpool destroy foobar
# newfs /dev/rdsk/c1t0d0s0
newfs: construct a new file system /dev/rdsk/c1t0d0s0: (y/n)? y
The device sector size 4096 is not supported by ufs!

John
groenveld at acm.org

Yeah, okay, duh. I should have known that large sector size
support is only available for a non-root ZFS file system.

A couple more things if you''re still interested:

1. If you re-create the pool on the whole disk, like this:

# zpool create foo c1t0d0

Then, resend the prtvtoc output for c1t0d0s0.

We should be able to tell if format is creating a dummy label,
which means the ZFS data is never getting written to this disk.
This would be a bug.

2. You are running this early S11 release:

SunOS 5.11 151.0.1.12 i386

You might retry this on more recent bits, like the EA release,
which I think is b 171.

I''ll still file the CR.

Thanks,

Cindy





On 10/13/11 09:40, John D Groenveld wrote:
> In message <201110131150.p9DBo8Yk011167 at acsinet22.oracle.com>,
Casper.Dik at oracl
> e.com writes:
>> What is the partition table?
> 
> I thought about that so I reproduced with the legacy SMI label
> and a Solaris fdisk partition with ZFS on slice 0.
> Same result as EFI; once I export the pool I cannot import it.
> 
> John
> groenveld at acm.org
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <4E9DA8B1.7020302 at oracle.com>, Cindy Swearingen
writes:
>1. If you re-create the pool on the whole disk, like this:
>
># zpool create foo c1t0d0
>
>Then, resend the prtvtoc output for c1t0d0s0.
# zpool create snafu c1t0d0
# zpool status snafu
  pool: snafu
 state: ONLINE
 scan: none requested
config:

        NAME        STATE     READ WRITE CKSUM
        snafu       ONLINE       0     0     0
          c1t0d0    ONLINE       0     0     0

errors: No known data errors
# prtvtoc /dev/rdsk/c1t0d0s0
* /dev/rdsk/c1t0d0s0 partition map
*
* Dimensions:
*    4096 bytes/sector
* 732566642 sectors
* 732566631 accessible sectors
*
* Flags:
*   1: unmountable
*  10: read-only
*
* Unallocated space:
*       First     Sector    Last
*       Sector     Count    Sector
*           6       250       255
*
*                          First     Sector    Last
* Partition  Tag  Flags    Sector     Count    Sector  Mount Directory
       0      4    00        256 732549997 732550252
       8     11    00  732550253     16384 732566636
>We should be able to tell if format is creating a dummy label,
>which means the ZFS data is never getting written to this disk.
>This would be a bug.
# zdb -l /dev/dsk/c1t0d0s0
--------------------------------------------
LABEL 0
--------------------------------------------
failed to unpack label 0
--------------------------------------------
LABEL 1
--------------------------------------------
failed to unpack label 1
--------------------------------------------
LABEL 2
--------------------------------------------
failed to unpack label 2
--------------------------------------------
LABEL 3
--------------------------------------------
failed to unpack label 3
>2. You are running this early S11 release:
>
>SunOS 5.11 151.0.1.12 i386
>
>You might retry this on more recent bits, like the EA release,
>which I think is b 171.
Doubtful I''ll find time to install EA before S11 FCS''s
November launch.
>I''ll still file the CR.
Thank you.

John
groenveld at acm.org

This is CR 7102272.

cs

On 10/18/11 10:50, John D Groenveld wrote:
> In message <4E9DA8B1.7020302 at oracle.com>, Cindy Swearingen writes:
>> 1. If you re-create the pool on the whole disk, like this:
>>
>> # zpool create foo c1t0d0
>>
>> Then, resend the prtvtoc output for c1t0d0s0.
> 
> # zpool create snafu c1t0d0
> # zpool status snafu
>   pool: snafu
>  state: ONLINE
>  scan: none requested
> config:
> 
>         NAME        STATE     READ WRITE CKSUM
>         snafu       ONLINE       0     0     0
>           c1t0d0    ONLINE       0     0     0
> 
> errors: No known data errors
> # prtvtoc /dev/rdsk/c1t0d0s0
> * /dev/rdsk/c1t0d0s0 partition map
> *
> * Dimensions:
> *    4096 bytes/sector
> * 732566642 sectors
> * 732566631 accessible sectors
> *
> * Flags:
> *   1: unmountable
> *  10: read-only
> *
> * Unallocated space:
> *       First     Sector    Last
> *       Sector     Count    Sector
> *           6       250       255
> *
> *                          First     Sector    Last
> * Partition  Tag  Flags    Sector     Count    Sector  Mount Directory
>        0      4    00        256 732549997 732550252
>        8     11    00  732550253     16384 732566636
> 
>> We should be able to tell if format is creating a dummy label,
>> which means the ZFS data is never getting written to this disk.
>> This would be a bug.
> 
> # zdb -l /dev/dsk/c1t0d0s0
> --------------------------------------------
> LABEL 0
> --------------------------------------------
> failed to unpack label 0
> --------------------------------------------
> LABEL 1
> --------------------------------------------
> failed to unpack label 1
> --------------------------------------------
> LABEL 2
> --------------------------------------------
> failed to unpack label 2
> --------------------------------------------
> LABEL 3
> --------------------------------------------
> failed to unpack label 3
> 
>> 2. You are running this early S11 release:
>>
>> SunOS 5.11 151.0.1.12 i386
>>
>> You might retry this on more recent bits, like the EA release,
>> which I think is b 171.
> 
> Doubtful I''ll find time to install EA before S11 FCS''s
> November launch.
> 
>> I''ll still file the CR.
> 
> Thank you.
> 
> John
> groenveld at acm.org
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

In message <4E9DB04B.80506 at oracle.com>, Cindy Swearingen
writes:
>This is CR 7102272.
Anyone out there have Western Digital''s competing 3TB Passport
drive handy to duplicate this bug?

John
groenveld at acm.org

In message <4E9DB04B.80506 at oracle.com>, Cindy Swearingen
writes:
>This is CR 7102272.
What is the title of this BugId?
I''m trying to attach my Oracle CSI to it but Chuck Rozwat
and company''s support engineer can''t seem to find it.

Once I get upgraded from S11x SRU12 to S11, I''ll reproduce
on a more recent kernel build.

Thanks,
John
groenveld at acm.org

Hi John,

CR 7102272:

  ZFS storage pool created on a 3 TB USB 3.0 device has device label 
problems

Let us know if this is still a problem in the OS11 FCS release.

Thanks,

Cindy


On 11/10/11 08:55, John D Groenveld wrote:
> In message<4E9DB04B.80506 at oracle.com>, Cindy Swearingen writes:
>> This is CR 7102272.
>
> What is the title of this BugId?
> I''m trying to attach my Oracle CSI to it but Chuck Rozwat
> and company''s support engineer can''t seem to find it.
>
> Once I get upgraded from S11x SRU12 to S11, I''ll reproduce
> on a more recent kernel build.
>
> Thanks,
> John
> groenveld at acm.org
>
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss

On Tue, Oct 11, 2011 at 08:17:55PM -0400, John D Groenveld
wrote:
> Under both Solaris 10 and Solaris 11x, I receive the evil message:
> | I/O request is not aligned with 4096 disk sector size.
> | It is handled through Read Modify Write but the performance is very low.
I got similar with 4k sector ''disks'' (as a comstar target with
blk=4096) when trying to use them to force a pool to ashift=12. The
labels are found at the wrong offset when the block numbers change,
and maybe the GPT label has issues too. 

--
Dan.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20111111/0405f1e0/attachment.bin>

On Nov 10, 2011, at 18:41, Daniel Carosone wrote:
> On Tue, Oct 11, 2011 at 08:17:55PM -0400, John D Groenveld wrote:
>> Under both Solaris 10 and Solaris 11x, I receive the evil message:
>> | I/O request is not aligned with 4096 disk sector size.
>> | It is handled through Read Modify Write but the performance is very
low.
> 
> I got similar with 4k sector ''disks'' (as a comstar target
with
> blk=4096) when trying to use them to force a pool to ashift=12. The
> labels are found at the wrong offset when the block numbers change,
> and maybe the GPT label has issues too. 
Anyone know if Solaris 11 has better support for detecting the native block size
of the underlying storage?

PSARC 2008/769 ("Multiple disk sector size support") was committed in
OpenSolaris in commit revision  9889:68d0fe4c716e. It appears ZFS makes use of
the check when opening a vdev:

http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/vdev_disk.c#287

Has anyone had a chance to play with S11 to confirm?

We''re only going to get more and more Advanced Format drives, never
mind all the SAN storage units out there as well (and VMFS often on top of that
too).

On Nov 10, 2011, at 7:47 PM, David Magda wrote:
> On Nov 10, 2011, at 18:41, Daniel Carosone wrote:
> 
>> On Tue, Oct 11, 2011 at 08:17:55PM -0400, John D Groenveld wrote:
>>> Under both Solaris 10 and Solaris 11x, I receive the evil message:
>>> | I/O request is not aligned with 4096 disk sector size.
>>> | It is handled through Read Modify Write but the performance is
very low.
>> 
>> I got similar with 4k sector ''disks'' (as a comstar
target with
>> blk=4096) when trying to use them to force a pool to ashift=12. The
>> labels are found at the wrong offset when the block numbers change,
>> and maybe the GPT label has issues too. 
> 
> Anyone know if Solaris 11 has better support for detecting the native block
size of the underlying storage?
Better than ?
If the disks advertise 512 bytes, the only way around it is with a whitelist. I
would
be rather surprised if Oracle sells 4KB sector disks for Solaris systems?
 -- richard

-- 

ZFS and performance consulting
http://www.RichardElling.com
LISA ''11, Boston, MA, December 4-9

On Fri, Nov 11, 2011 at 09:55:29PM -0800, Richard Elling
wrote:
> On Nov 10, 2011, at 7:47 PM, David Magda wrote:
> 
> > On Nov 10, 2011, at 18:41, Daniel Carosone wrote:
> > 
> >> On Tue, Oct 11, 2011 at 08:17:55PM -0400, John D Groenveld wrote:
> >>> Under both Solaris 10 and Solaris 11x, I receive the evil
message:
> >>> | I/O request is not aligned with 4096 disk sector size.
> >>> | It is handled through Read Modify Write but the performance
is very low.
> >> 
> >> I got similar with 4k sector ''disks'' (as a
comstar target with
> >> blk=4096) when trying to use them to force a pool to ashift=12.
The
> >> labels are found at the wrong offset when the block numbers
change,
> >> and maybe the GPT label has issues too. 
> > 
> > Anyone know if Solaris 11 has better support for detecting the native
block size of the underlying storage?
> 
> Better than ?
> If the disks advertise 512 bytes, the only way around it is with a
whitelist. I would
> be rather surprised if Oracle sells 4KB sector disks for Solaris systems?
>
Afaik the disks advertise both the physical and logical sector size..
at least on Linux you can see that the disk emulates 512 bytes/sector,
but natively it uses 4kB/sector.

/sys/block/<disk>/queue/logical_block_size=512
/sys/block/<disk>/queue/physical_block_size=4096

The info should be available through IDENTIFY DEVICE (ATA) or READ CAPACITY 16
(SCSI) commands.

-- Pasi

On Nov 12, 2011, at 00:55, Richard Elling wrote:
> Better than ?
> If the disks advertise 512 bytes, the only way around it is with a
whitelist. I would
> be rather surprised if Oracle sells 4KB sector disks for Solaris systems?
Solaris 10. OpenSolaris.

But would it be surprising to use SANs with Solaris? Or perhaps run Solaris
under some kind of virtualized environment where the virtual disk has a
particular block size? Or maybe SSDs, which tend to read/write/delete in certain
block sizes?

In these situations simply assuming 512 may slow things down.

And if Solaris 11 is going to be around for a decade or so, I''d hazard
to guess that 512B sector disks will become less and less prevalent as time goes
on. Might as well enable the functionality now, when 4K is rarer, so you have
more time to test and tunes things out?rather than later when you can
potentially be left scrambling.

As Pasi K?rkk?inen mentions, there''s not much you can do if the disks
lies (just as has been seen with disks that lie about flushing the cache). This
is mostly a temporary kludge for legacy''s sake. More and more disks
will be truthful as times goes on.

On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda
wrote:
> On Nov 12, 2011, at 00:55, Richard Elling wrote:
> 
> > Better than ?
> > If the disks advertise 512 bytes, the only way around it is with a
whitelist. I would
> > be rather surprised if Oracle sells 4KB sector disks for Solaris
systems?
> 
> Solaris 10. OpenSolaris.
> 
> But would it be surprising to use SANs with Solaris? Or perhaps run Solaris
under some kind of virtualized environment where the virtual disk has a
particular block size? Or maybe SSDs, which tend to read/write/delete in certain
block sizes?
> 
> In these situations simply assuming 512 may slow things down.
> 
> And if Solaris 11 is going to be around for a decade or so, I''d
hazard to guess that 512B sector disks will become less and less prevalent as
time goes on. Might as well enable the functionality now, when 4K is rarer, so
you have more time to test and tunes things out?rather than later when you can
potentially be left scrambling.
> 
> As Pasi K?rkk?inen mentions, there''s not much you can do if the
disks lies (just as has been seen with disks that lie about flushing the cache).
This is mostly a temporary kludge for legacy''s sake. More and more
disks will be truthful as times goes on.
> 
Most "4kB"/sector disks already today properly report both the
physical (4kB) and logical (512b) sector sizes.
It sounds like *solaris is only checking the logical (512b) sector size, not the
physical (4kB) sector size..

-- Pasi

On Nov 12, 2011, at 8:31 AM, Pasi K?rkk?inen wrote:
> On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda wrote:
>> On Nov 12, 2011, at 00:55, Richard Elling wrote:
>> 
>>> Better than ?
>>> If the disks advertise 512 bytes, the only way around it is with a
whitelist. I would
>>> be rather surprised if Oracle sells 4KB sector disks for Solaris
systems?
>> 
>> Solaris 10. OpenSolaris.
>> 
>> But would it be surprising to use SANs with Solaris? Or perhaps run
Solaris under some kind of virtualized environment where the virtual disk has a
particular block size? Or maybe SSDs, which tend to read/write/delete in certain
block sizes?
>> 
>> In these situations simply assuming 512 may slow things down.
>> 
>> And if Solaris 11 is going to be around for a decade or so,
I''d hazard to guess that 512B sector disks will become less and less
prevalent as time goes on. Might as well enable the functionality now, when 4K
is rarer, so you have more time to test and tunes things out?rather than later
when you can potentially be left scrambling.
>> 
>> As Pasi K?rkk?inen mentions, there''s not much you can do if
the disks lies (just as has been seen with disks that lie about flushing the
cache). This is mostly a temporary kludge for legacy''s sake. More and
more disks will be truthful as times goes on.
>> 
> 
> Most "4kB"/sector disks already today properly report both the
physical (4kB) and logical (512b) sector sizes.
> It sounds like *solaris is only checking the logical (512b) sector size,
not the physical (4kB) sector size..
ZFS uses the physical block size.
http://src.illumos.org/source/xref/illumos-gate/usr/src/uts/common/fs/zfs/vdev_disk.c#294

 -- richard

-- 

ZFS and performance consulting
http://www.RichardElling.com
LISA ''11, Boston, MA, December 4-9

On Sat, Nov 12, 2011 at 10:08:04AM -0800, Richard Elling
wrote:
> 
> On Nov 12, 2011, at 8:31 AM, Pasi K?rkk?inen wrote:
> 
> > On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda wrote:
> >> On Nov 12, 2011, at 00:55, Richard Elling wrote:
> >> 
> >>> Better than ?
> >>> If the disks advertise 512 bytes, the only way around it is
with a whitelist. I would
> >>> be rather surprised if Oracle sells 4KB sector disks for
Solaris systems?
> >> 
> >> Solaris 10. OpenSolaris.
> >> 
> >> But would it be surprising to use SANs with Solaris? Or perhaps
run Solaris under some kind of virtualized environment where the virtual disk
has a particular block size? Or maybe SSDs, which tend to read/write/delete in
certain block sizes?
> >> 
> >> In these situations simply assuming 512 may slow things down.
> >> 
> >> And if Solaris 11 is going to be around for a decade or so,
I''d hazard to guess that 512B sector disks will become less and less
prevalent as time goes on. Might as well enable the functionality now, when 4K
is rarer, so you have more time to test and tunes things out?rather than later
when you can potentially be left scrambling.
> >> 
> >> As Pasi K?rkk?inen mentions, there''s not much you can do
if the disks lies (just as has been seen with disks that lie about flushing the
cache). This is mostly a temporary kludge for legacy''s sake. More and
more disks will be truthful as times goes on.
> >> 
> > 
> > Most "4kB"/sector disks already today properly report both
the physical (4kB) and logical (512b) sector sizes.
> > It sounds like *solaris is only checking the logical (512b) sector
size, not the physical (4kB) sector size..
> 
> ZFS uses the physical block size.
>
http://src.illumos.org/source/xref/illumos-gate/usr/src/uts/common/fs/zfs/vdev_disk.c#294
> 
Hmm.. so everything should just work? 
Does some other part of the code use logical block size then, for example to
calculate the ashift?

Maybe I should read the code :) 

-- Pasi

tip below?

On Nov 13, 2011, at 3:24 AM, Pasi K?rkk?inen wrote:
> On Sat, Nov 12, 2011 at 10:08:04AM -0800, Richard Elling wrote:
>> 
>> On Nov 12, 2011, at 8:31 AM, Pasi K?rkk?inen wrote:
>> 
>>> On Sat, Nov 12, 2011 at 08:15:31AM -0500, David Magda wrote:
>>>> On Nov 12, 2011, at 00:55, Richard Elling wrote:
>>>> 
>>>>> Better than ?
>>>>> If the disks advertise 512 bytes, the only way around it is
with a whitelist. I would
>>>>> be rather surprised if Oracle sells 4KB sector disks for
Solaris systems?
>>>> 
>>>> Solaris 10. OpenSolaris.
>>>> 
>>>> But would it be surprising to use SANs with Solaris? Or perhaps
run Solaris under some kind of virtualized environment where the virtual disk
has a particular block size? Or maybe SSDs, which tend to read/write/delete in
certain block sizes?
>>>> 
>>>> In these situations simply assuming 512 may slow things down.
>>>> 
>>>> And if Solaris 11 is going to be around for a decade or so,
I''d hazard to guess that 512B sector disks will become less and less
prevalent as time goes on. Might as well enable the functionality now, when 4K
is rarer, so you have more time to test and tunes things out?rather than later
when you can potentially be left scrambling.
>>>> 
>>>> As Pasi K?rkk?inen mentions, there''s not much you can
do if the disks lies (just as has been seen with disks that lie about flushing
the cache). This is mostly a temporary kludge for legacy''s sake. More
and more disks will be truthful as times goes on.
>>>> 
>>> 
>>> Most "4kB"/sector disks already today properly report
both the physical (4kB) and logical (512b) sector sizes.
>>> It sounds like *solaris is only checking the logical (512b) sector
size, not the physical (4kB) sector size..
>> 
>> ZFS uses the physical block size.
>>
http://src.illumos.org/source/xref/illumos-gate/usr/src/uts/common/fs/zfs/vdev_disk.c#294
>> 
> 
> Hmm.. so everything should just work? 
> Does some other part of the code use logical block size then, for example
to calculate the ashift?
> 
> Maybe I should read the code :) 
Or look at what your system reports :-) Though not directly intended for this
use,
	echo ::sd_state | mdb -k
and look for the un_phy_blocksize.
 -- richard

-- 

ZFS and performance consulting
http://www.RichardElling.com
LISA ''11, Boston, MA, December 4-9

In message <4EBBFB52.6 at oracle.com>, Cindy Swearingen
writes:
>CR 7102272:
>
>  ZFS storage pool created on a 3 TB USB 3.0 device has device label 
>problems
>
>Let us know if this is still a problem in the OS11 FCS release.
I finally got upgraded from Solaris 11 Express SRU 12 to S11 FCS.

Solaris 11 11/11 still spews the "I/O request is not aligned with
4096 disk sector size" warnings but zpool(1M) create''s label
persists and I can export and import between systems.

John
groenveld at acm.org