zfs discuss - Nov 2011 - After update to S11, zfs reports some disks as 'corrupted data'

Hi all,

I am in the process of updating my SE11 servers to S11. On one server I 
am having two zpools made out of mirror vdevs and up to now none of 
these zpools have shown any error.

However, after updating to S11 three disks - and unfortuanetly two of 
the same miorror vdev are shown as UNAVAIL due to ''corrupted
data'':

   pool: obelixData
     id: 9610325806378085846
  state: UNAVAIL
status: One or more devices contains corrupted data.
action: The pool cannot be imported due to damaged devices or data.
    see: http://www.sun.com/msg/ZFS-8000-5E
config:

         obelixData                  UNAVAIL  insufficient replicas
           mirror-0                  DEGRADED
             c9t2100001378AC02DDd0   ONLINE
             c9t2100001378AC02F4d0   UNAVAIL  corrupted data
           mirror-1                  UNAVAIL  insufficient replicas
             c9t2100001378AC02DDd1   UNAVAIL  corrupted data
             c9t2100001378AC02F4d1   UNAVAIL  corrupted data
           mirror-2                  ONLINE
             c9t2100001378AC02DDd2   ONLINE
             c9t2100001378AC02F4d2   ONLINE
           mirror-3                  ONLINE
             c9t2100001378AC02DDd3   ONLINE
             c9t2100001378AC02F4d3   ONLINE
           mirror-4                  ONLINE
             c9t2100001378AC02DDd5   ONLINE
             c9t2100001378AC02F4d5   ONLINE
           mirror-5                  ONLINE
             c9t2100001378AC02DDd4   ONLINE
             c9t2100001378AC02F4d4   ONLINE
           mirror-6                  ONLINE
             c9t2100001378AC02DDd6   ONLINE
             c9t2100001378AC02F4d6   ONLINE
           mirror-7                  ONLINE
             c9t2100001378AC02DDd7   ONLINE
             c9t2100001378AC02F4d7   ONLINE
           mirror-8                  ONLINE
             c9t2100001378AC02DDd8   ONLINE
             c9t2100001378AC02F4d8   ONLINE
           mirror-9                  ONLINE
             c9t2100001378AC02DDd9   ONLINE
             c9t2100001378AC02F4d9   ONLINE
           mirror-10                 ONLINE
             c9t2100001378AC02DDd10  ONLINE
             c9t2100001378AC02F4d10  ONLINE
           mirror-11                 ONLINE
             c9t2100001378AC02DDd11  ONLINE
             c9t2100001378AC02F4d11  ONLINE
           mirror-12                 ONLINE
             c9t2100001378AC02DDd12  ONLINE
             c9t2100001378AC02F4d12  ONLINE
           mirror-13                 ONLINE
             c9t2100001378AC02DDd13  ONLINE
             c9t2100001378AC02F4d13  ONLINE
           mirror-14                 ONLINE
             c9t2100001378AC02DDd14  ONLINE
             c9t2100001378AC02F4d14  ONLINE
         logs
           mirror-15                 ONLINE
             c9t2100001378AC02D9d0   ONLINE
             c9t2100001378AC02BFd0   ONLINE

So I reverted back to my saved SE11 BE and SE11 shows these disks as 
online and imnports this zpool without any issue.
I have now a scrub running on this zpool, but I am curious, if anyone 
had ever experienced something similar.

Thanks,
budy
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20111119/c1491d3a/attachment.html>

Phew? seems that the S11 update process replaced my modified qlc.conf 
with a standard one. In SE11 I had to lower the queue depth by setting 
max_execution_throttle to something lower than 16. Mostly since I am 
exposing 16 LUNs from each storage and this the qlc.driver flooded the 
storage controllers with I/Os.

Cheers,
budy