GTM health monitor fails on pool, but not on member.

my apologies if the following is overkill - just want to lay out some iquery functionality that is often vague in documentation and misunderstood until you have to dig in:

regarding the earlier comment "This GTM does communicate with LTMs using iquery, however the GTM does not use that data. All server/member/node objects in the GTM are added manually."

iquery is used between GTM/LTM not only to autodiscover LTM virtual servers and monitor health of the F5 devices, but also to pass on health status of virtual servers from the LTM on which a vs is configured to a GTM that is configured with that vs as a GTM vs object, regardless of whether the server object was manually added or autodiscovered - if a GTM vs was manually added to the LTM server object and a bigip monitor is assigned to that server object, the GTM will ascertain the vs status via its iquery connection from that LTM.

when a specific non-bigip monitor is assigned to a GTM pool/member/server, the GTM may delegate that monitoring activity to a different iquery-connected F5-device. it doesn't necessarily handle the monitoring itself. from my experience, it will choose a live F5 device (GTM or LTM) in the same GTM data center as the monitored virtual-server object. for example, if an http monitor is assigned to a GTM pool/member in DC1, the GTM may tell another LTM in DC1 to monitor that object and relay the status back via iquery. this can cause issues if there's a firewall that prevents the selected F5 device from hitting that server object via port 80, 443, etc. required by the non-bigip monitor