Probe should distinguish error from SP and error from IdP

[alexstuart]

Using the UK federation Test IdP in the list of IdPs to probe but removing that IdP's metadata from the SP leads to

2019-06-25 07:38:13 WARN Shibboleth.SessionInitiator.SAML2 [4] [default]: unable to locate metadata for provider (https://test-idp.ukfederation.org.uk/idp/shibboleth)

Probe doesn't recognise this is a problem with the SP and reports lines like

2019-06-25T07:35:44Z, https://test-idp.ukfederation.org.uk/idp/shibboleth, 500

which are indistinguishable from results where the metadata hasn't propagated to the IdP yet.

[alexstuart]

Solution should be something like a pre-flight check using a modification of the probe command . Something like curl --connect-timeout 3 -k --silent -c /tmp/cookies.txt ' . $RequestInitiator . '?entityID=' . $encodedID (not following any redirects) which will give HTTP 500 error if the SP fails to respond.