This looks good, but I'd prefer we put a small comment in as to why we are doing the additional wait. A summary of what you've said above for future reference in the test. Namely, upgrades occur on agent and potentially db as well. Juju doesn't tell us when it's truly done. We are using list-models as the proxy for 'juju done'.
This looks good, but I'd prefer we put a small comment in as to why we are doing the additional wait. A summary of what you've said above for future reference in the test. Namely, upgrades occur on agent and potentially db as well. Juju doesn't tell us when it's truly done. We are using list-models as the proxy for 'juju done'.