update icq healthcheck

[baabeetaa]

• qsd_ibc_channel_state - 0 = init, 1 = open, 2 = closed - if it's not 1 we have issues (0 means channel has been reopened and hermes is not moving channel handshakes about; 2 means channel is closed and needs reopening (and relayers configuring with new channels)
• qsd_ibc_commitments and qsd_ibc_acks - 0 = no packets/acks pending. non-zero and not decreasing == relaying issues
• qsd_icq_oldest_emission_distance - the age of the current oldest icq request. this should never really be over 200 in normal operation
• qsd_icq_historic_queue - size of icq unanswered queries queue - should generally be below 100 (and mostly below 10). if persistently above this, there is some issue. (edited)

[baabeetaa]

if packets are not shipped, channels will timeout and close after six hours
qsd_ibc_commitments and qsd_ibc_acks are the appropriate metrics; they should always return to zero (sometimes peak at around 100 before returning back down to zero)

qsd_icq_oldest_emission_distance is probably the surest indicator something is wrong. many icq responses will fail if the delegate channel is closed.
if this starts growing 500+ there is some issue (either funds, rpc down or channel closure)