Looking at the system logs during an installation, I think what we're seeing is that when curtin completes an install, it logs out of the iscsi connection but hasn't flushed the io devices completely;  in this case ,there's an lvm over the iscsi disk and the logout happens before everything is umounted.

Apr 20 20:23:13 civil-cattle cloud-init[3057]: Running command ['sync'] with allowed return codes [0] (shell=False, capture=False)
Apr 20 20:23:13 civil-cattle cloud-init[3057]: iscsiadm_logout: target=iqn.maas.io:priyank portal=10.1.0.100:3260
Apr 20 20:23:13 civil-cattle cloud-init[3057]: Running command ['iscsiadm', '--mode=node', '--targetname=iqn.maas.io:priyank', '--portal=10.1.0.100:3260', '--logout'] with allowed return codes [0] (shell=False, capture=True)
Apr 20 20:23:13 civil-cattle kernel: [  195.813504] sd 11:0:0:1: [sdc] Synchronizing SCSI cache
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Command returned stdout=Logging out of session [sid: 2, target: iqn.maas.io:priyank, portal: 10.1.0.100,3260]
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Logout of [sid: 2, target: iqn.maas.io:priyank, portal: 10.1.0.100,3260] successful.
Apr 20 20:23:14 civil-cattle cloud-init[3057]: , stderr=
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Running command ['udevadm', 'settle'] with allowed return codes [0] (shell=False, capture=False)
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Running command ['lsblk', '--noheadings', '--bytes', '--pairs', '--output=ALIGNMENT,DISC-ALN,DISC-GRAN,DISC-MAX,DISC-ZERO,FSTYPE,GROUP,KNAME,LABEL,LOG-SEC,MAJ:MIN,MIN-IO,MODE,MODEL,MOUNTPOINT,NAME,OPT-IO,OWNER,PHY-SEC,RM,RO,ROTA,RQ-SIZE,SIZE,STATE,TYPE,UUID'] with allowed return codes [0] (shell=False, capture=True)
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Running command ['umount', '/tmp/tmplwixwl87/target/mnt'] with allowed return codes [0] (shell=False, capture=False)
Apr 20 20:23:14 civil-cattle iscsid: Connection2:0 to [target: iqn.maas.io:priyank, portal: 10.1.0.100,3260] through [iface: default] is shutdown.
Apr 20 20:23:14 civil-cattle kernel: [  196.375256] Buffer I/O error on dev dm-0, logical block 1081344, lost sync page write
Apr 20 20:23:14 civil-cattle kernel: [  196.392268] JBD2: Error -5 detected when updating journal superblock for dm-0-8.
Apr 20 20:23:14 civil-cattle kernel: [  196.407032] Aborting journal on device dm-0-8.
Apr 20 20:23:14 civil-cattle kernel: [  196.415902] Buffer I/O error on dev dm-0, logical block 1081344, lost sync page write
Apr 20 20:23:14 civil-cattle kernel: [  196.431540] JBD2: Error -5 detected when updating journal superblock for dm-0-8.
Apr 20 20:23:14 civil-cattle kernel: [  196.446326] Buffer I/O error on dev dm-0, logical block 0, lost sync page write
Apr 20 20:23:14 civil-cattle kernel: [  196.460917] EXT4-fs error (device dm-0): ext4_put_super:813: Couldn't clean up the journal
Apr 20 20:23:14 civil-cattle kernel: [  196.477463] EXT4-fs (dm-0): Remounting filesystem read-only
Apr 20 20:23:14 civil-cattle kernel: [  196.488581] EXT4-fs (dm-0): previous I/O error to superblock detected
Apr 20 20:23:14 civil-cattle kernel: [  196.501446] Buffer I/O error on dev dm-0, logical block 0, lost sync page write
Apr 20 20:23:14 civil-cattle cloud-init[3057]: Running command ['umount', '/tmp/tmplwixwl87/target'] with allowed return codes [0] (shell=False, capture=False)
Apr 20 20:23:14 civil-cattle cloud-init[3057]: finish: cmd-install: SUCCESS: curtin command install