Li, This seems OK to me. On x86, the first cylinder is reserved, so the starting sector is moved to the second cylinder. The partition table you list below looks fine to me.
Run format->fdisk on the disk to check the partition. FYI, partition_start_sector is optional, and not declaring a partition_start_sector in a partition_action=create would cause the new partition to be placed at the front (the front of whatever free region on the disk fits best). Also, partition_number is ignored for partition_action=create - only used for partition_action=delete. William Li Fan wrote: > Hi, William, > > I am trying to verify that if partition_start_sector is set to 0, > automated installation will fail. > > I created the following manifest, > > <ai_manifest name="ai_test_manifest"> > <ai_target_device> > <target_device_name>c7t0d0p0</target_device_name> > > <target_device_use_solaris_partition>false</target_device_use_solaris_partition> > > > > <target_device_overwrite_root_zfs_pool>true</target_device_overwrite_root_zfs_pool> > > > </ai_target_device> > <ai_device_partitioning> > <partition_action>create</partition_action> > <partition_number>1</partition_number> > <partition_start_sector>0</partition_start_sector> > <partition_size>20480</partition_size> > <partition_type>191</partition_type> > </ai_device_partitioning> > <ai_pkg_repo_default_authority> > <main url="IPS_URL" authname="AUTHNAME"/> > <mirror url=""/> > </ai_pkg_repo_default_authority> > <ai_packages> > </ai_packages> > </ai_manifest> > > > and use the option "-I" of auto-install. > The manifest passed validation. After target initiation was completed, > following fdisk table was got: > > * Id Act Bhead Bsect Bcyl Ehead Esect Ecyl Rsect > Numsect > 191 0 0 1 1 254 63 1023 16065 > 41926975 > The actual start sector was set to 16065. > > I was using osol-0906-110-ai-x86.iso image. > > > > Thanks > --Li Fan > > > > >
