start_server_for_group may hit internal error on shutdown about scylladb

[bhalevy]

Assuming it's not a blocker

We can live with it for the time being since it's on the shutdown path and causes a crash in testing (though I'm not sure about the outcome if we don't abort on this internal error)

[kbr-scylla]

Assuming it's not a blocker

[mykaul]

Assuming it's not a blocker

We can live with it for the time being since it's on the shutdown path and causes a crash in testing (though I'm not sure about the outcome if we don't abort on this internal error)

Is it on the shutdown path?
Another failure with dtest-debug @ https://jenkins.scylladb.com/job/scylla-enterprise/job/dtest-debug/163/artifact/logs-full.debug.003/1714891843992_update_cluster_layout_tests.py%3A%3ATestStopNodeEarly%3A%3Atest_stop_node_while_restarting%5BTrue-False%5D/

Node log attached.
18197_node1.log

[mykaul]

And here (same run, different test) - https://jenkins.scylladb.com/job/scylla-enterprise/job/dtest-debug/163/artifact/logs-full.debug.002/1714891856631_update_cluster_layout_tests.py%3A%3ATestStopNodeEarly%3A%3Atest_stop_node_while_restarting%5BTrue-False%5D/

[patjed41]

I've investigated #17794 before realizing it's a dup of this issue, which is not considered a release blocker. Since the fix is trivial, we've decided to fix it anyway: #18569.