ctl.conf / iscsi docs and best practices

Looking at diving deeper into FreeBSD as an iscsi target and initiator, 
but more so as a target server.  Looking at the handbook, there are not 
a lot of docs there and the sample ctl.conf entries dont really follow 
the style of the example ctl.conf file generated by something like 
TrueNAS.  Are there any better FreeBSD docs out there ? Anyone have any 
decent go to docs around iscsi in general that are good primers as well 
? My goal is to serve up storage for a bunch of Linux KVM initiators 
with the iscsi targets on FreeBSD being zvols.

e.g. in TrueNAS, I see it define a bunch of individual LUNs and then 
targets as so.

lun "freebsd2" {
         ctl-lun "1"
         path "/dev/zvol/tankstripe/freebsd2"
         blocksize "512"
         serial "000743299890001"
         device-id "iSCSI Disk      000743299890001 "
         option "vendor" "TrueNAS"
         option "product" "iSCSI Disk"
         option "revision" "0123"
         option "naa" "0x6589cfc000000c330e3ddde42a36408a"
         option "insecure_tpc" "on"
         option "rpm" "1"
}

lun "freebsd3" {
         ctl-lun "2"
         path "/dev/zvol/tankstripe/freebsd3"
         blocksize "4096"
         serial "000743299890002"
         device-id "iSCSI Disk      000743299890002 "
         option "vendor" "TrueNAS"
         option "product" "iSCSI Disk"
         option "revision" "0123"
         option "naa" "0x6589cfc0000001cd46c021f21b31d0cc"
         option "insecure_tpc" "on"
         option "rpm" "1"
}

Then define more LUNs in the targets that reference the above LUNs

target "iqn.2005-10.org.freenas.ctl:freebsd2" {
         alias "freebsd2"
         portal-group "pg1" "ag4tg2_2"

         lun "0" "freebsd2"
}

target "iqn.2005-10.org.freenas.ctl:freebsd3" {
         alias "freebsd3"
         portal-group "pg1" "ag4tg3_3"

         lun "0" "freebsd3"
}


So the lun that refers to the block of storage also has a ctl-lun that 
seems to be uniq. What is the purpose of that ? Same with serial #s and 
Device-IDs. Are these merely informational or are they a requirement for 
something ?

If multiple initiators are to have sessions to the target, with only one 
actively writing to it, are there any special options I need to set ?

I also noticed in ctladm are several cache sync options.  My go - to zfs 
replication software is zrepl which has hooks that can be fired before 
and after a snapshot.  Do I want to send some sort of cache flush 
command to the target ? Or is that the guest OS' responsibility to do a 
sync that that the snapshot is coherent as possible.

Thanks for any pointers.

     ---Mike