Orderer check via logs is not robust

[nicolapaoli]

After deploying a network and leaving it for some time without activity, the orderer log that check it is functioning stalls.

[inzamam-iqbal]

Hyperledger 1.4 peer and orderer supports helath checks where we can just get the health status by sending a GET request to /healthz. So utilizing that feature after updating to hyperladger 1.4 will be the best solution for this.

[alexvicegrab]

Agreed, I would suggest to add these health checks, and if we don't find the /healthz, to default to the old code, but add a deprecation notice. This would mention that in 3/6/12 months (I favour 6, but we can perhaps vote on this?), we will stop supporting Fabric >1.4, so that this change does not come as a surprise.