Allow 307 return code

[irivera007]

When deployed the Redis container keeps crashing given that the services cannot get a healthy check, allowing 307 fixes this.

[petr0sec]

I was about to open a PR for this exact thing. Wish I would have found this before spending an hour troubleshooting why my containers were crash looping. Thanks for opening this @irivera007

[petr0sec]

FWIW, 307 is really not a proper response code for a health check. I am guessing they changed the healthcheck endpoint (possibly to /app ?) and that's probably what needs to be updated. I'll let 1password confirm though. FWIW, I've changed the healthcheck endpoint to /app on my end rather than the status code and it seems to be stable now.

[irivera007]

@petr0sec thank you!
your approach makes more sense, thank you for sharing. I agree lets see what 1pwd says about the final solution for this. Im sure this benefits the community as you saw it could be real struggle troubleshooting

[devilleweppenaar]

Thank you for creating this PR and providing details. We're looking into this and will provide feedback shortly.

In the meantime, since release 2.3.0 of the SCIM bridge you can now enable an optional ping server that can be used for health check purposes. Some more details are available in the linked changelog or the help output of the SCIM bridge (./op-scim --help).

[chasdl]

Thanks for all the feedback and help here. This was an oversight in our latest release.

Prior to release v2.3.0 of the SCIM bridge, hitting GET / would return a 200 if everything was healthy, and therefore was a proper path for health checks. Now with v2.3.0 we are redirecting / to /app in order to serve the new React app, so defining /app as the health check path by default gets us to a working state.

As @devillexio mentioned, since release v2.3.0, you can enable an optional ping server on /ping for health check purposes. This will be the way forward as the default for our helm charts, but as this requires the OP_PING_SERVER option to be enabled and is disabled by default, we can stick to /app for this config.

Changes made in #162