Every loop over a list of names ran the xCAT clients on standard input, so the first client read the rest of the file and the loop ended after one name. Teardown said it had removed the fixture while five nodes, an osimage, a network and an unrestored site table were still there, and the next setup recorded those leftovers as the state to restore to. The loops now read on descriptor 3. `nodeset offline` is not enough either: it gives up when it cannot reach the DHCP backend and never removes the staged kernel and initrd. The artefacts the fixture's own nodes and images can produce are now removed by name. check refuses on any leftover node, not just the first.
xCAT-test
Unit tests that run from the source checkout are split by implementation language:
| Test type | Location | Runner |
|---|---|---|
| Perl unit tests | xCAT-test/unit/*.t |
prove -r xCAT-test/unit |
| Shell unit tests | xCAT-test/bats/*.bats |
bats -r xCAT-test/bats |
| DHCP wire tests | xCAT-test/dhcptest/ |
python3 -m unittest discover -s tests, then dhcptest run -i <nic> <conf> |
| CLI functional tests | xCAT-test/autotest/testcase/ and xCAT-test/autotest/bundle/ |
xcattest -f <cluster.conf> -t <case> or xcattest -f <cluster.conf> -b <bundle> |
Use Perl .t tests for Perl modules, Perl scripts, templates, and repository
artifacts. Use BATS tests for shell-script behavior that can be exercised from
the checkout by sourcing a shell library or script and shadowing external
commands.
Shell behavior should not be tested by Perl tests that grep shell source. Put
those tests under xCAT-test/bats instead.
dhcptest is the odd one out: its Python unit tests and dhcptest validate
run offline like the others, but its point is the third column -- driving real
DHCP transactions on a provisioning NIC and asserting on what came back. It
talks to the server only over the wire, so it is agnostic to which DHCP
implementation is answering and to xCAT itself.
See unit/README.md, bats/README.md and dhcptest/README.md for the
detailed rules for each suite.