I have following setup: prez behind nginx in one kubernetes pod/vm, each in separate container but on the same deployment.
I have no control which starts first, though
Out of the box kubernetes healthcheck probes are based on HTTP GET which I configures to probe '/' of prez
But it looks like prez fails to respond on http GET '/' when SYSTEM_URI is not ready yet.
I can play with delays, custom probes etc but it all remains quite fragile to have basic probe dependent on the other containers. An option would be sth like Fuseki /ping which echoes date.
I have following setup: prez behind nginx in one kubernetes pod/vm, each in separate container but on the same deployment.
I have no control which starts first, though
Out of the box kubernetes healthcheck probes are based on HTTP GET which I configures to probe '/' of prez
But it looks like prez fails to respond on http GET '/' when SYSTEM_URI is not ready yet.
I can play with delays, custom probes etc but it all remains quite fragile to have basic probe dependent on the other containers. An option would be sth like Fuseki /ping which echoes date.