When trying to connect to an additional iSCSI target/server with the widely
reported command
   iscsiadm -m node --target iqn.foo --login
iscsiadm on at least some networks will fail with incorrect error messages:
   Logging in to [iface: ifBar_0, target: iqn.foo, portal: 10.x.x.x,3260]
   iscsiadm: Could not login to [iface: ifBar_0, target: iqn.foo,
portal: 10.x.x.x,3260].
   iscsiadm: initiator reported error (19 - encountered non-retryable
iSCSI login failure)
   iscsiadm: Could not log into all portals

However running a simultaneous tcpdump and analyzing it with WireShark
reveals
that iscsiadm does a succesful login, then generates a TCP RST ending the
attempt for no apparent reason.

Because whatever failed was something local on the client, the error message
must say so and be much more actionable instead of sending the sysadmin on
a wild goose chase through the authentication settings and target
configuration.

P.S.
This was run against another bookworm machine running targetcli-fb, no
exotic
hardware needed.  All firewalls that I could find on the path are off.

P.P.S.
As stated in the first line of the report, this happens for an additional
iSCSI connection while the client already has iSCSI connections to other
disc boxes (in this case on other network interfaces).