[Xen-devel] window 2003 blue screen under IO stress test

[MaoXiaoyun <tinnycloud@xxxxxxxxxxx>]

Hi James:
 
          Recently we are continue more and more test on window HVM with PV driver installed.
          We start about 20 HVMs on a single host, and do IO stress test inside HVM.
          
          Well some of the VM hit blue screen in half an hour also.
          The BSD Code is: KERNEL_DATA_INPAGE_ERROR
          STOP: 0x0000007A(0x07B8C38, 0xC07B8C28, 0xF718771C, 0X36DF6860)
           or    0x0000007E (0x80000003, 0x80878113, 0xF7912254, 0xF7911F50)
 
          Attached logs is the debug output, you can find the BSD code at the End.
          Althrough two logs has different BSD code, but they have common abnormal logs like:
 
12955411748328: XenVbd     SCSIOP_MODE_SENSE llbaa = 0, dbd = 0, page_code = 63, allocation_length = 192
12955411748328: XenVbd     SCSIOP_MODE_SENSE llbaa = 0, dbd = 0, page_code = 63, allocation_length = 192
12955411752281: XenVbd     SCSIOP_MODE_SENSE llbaa = 0, dbd = 0, page_code = 63, allocation_length = 192
12955411752281: XenVbd     SCSIOP_MODE_SENSE llbaa = 0, dbd = 0, page_code = 63, allocation_length = 192

         and later many XenVbd_HwScsiResetBus.
 
         Could enlighten me what is going on in the VM, does VM read corrupted data
(since we have our on distributed storage), or due to the large IO response in heavy stress.
 
         Thanks for your help.
 
 
 

Attachment: qemu-dm-w2003-1.vhd.log.3
Description: Binary data

Attachment: qemu-dm-w2003-9.vhd.log.5
Description: Binary data

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel