[WPB-23644] refactor user type in wire-subsystems

[fisx]

Historically, we kept track of the user type "app" only in postgres. By now we have duplicated this information into cassandra, so a lot of code becomes more simple to write. This PR is fixing a few spots that I've missed when introducing user type storage in cassandra.

Checklist

• Add a new entry in an appropriate subdirectory of changelog.d
• Read and follow the PR guidelines

[akshaymankar]

I see you've checked the item for having read and followed PR guidelines, but I don't see an update to the v15 API. Perhaps this should also be part of changelog?

[fisx]

https://wearezeta.atlassian.net/wiki/spaces/ENGINEERIN/history/2570879022/API+changes+v14+-+v15

[akshaymankar]

Do you know why this is changing?

[fisx]

i think it's equivalent. copilot didn't like the previous one.

[fisx]

should i roll it back?

[akshaymankar]

Did copilot generate this JSON? Or is it generated by aeson?

[fisx]

i asked copilot to fix the golden tests. i also read the changes, just didn't bother to undo this one because i saw no harm.

i've changed it back: b77e7b8

in hindsight i should have just done something with find, and perl, would have been much faster both in implementation and in code review...

[akshaymankar]

I think it would be nice if these JSON files can be automatically regenerated and our current machinery is maintained for this. It shouldn't require extra massaging. I'll dig into what happened here, for now I guess its ok.

[akshaymankar]

Nits inline.

[akshaymankar]

I think it would be nice if these JSON files can be automatically regenerated and our current machinery is maintained for this. It shouldn't require extra massaging. I'll dig into what happened here, for now I guess its ok.

[fisx]

The CI errors consistently happen on this branch and nowhere else, but I don't understand:

STDERR:
  I0309 10:14:26.035245    2534 warnings.go:110] "Warning: would violate PodSecurity \"restricted:latest\": allowPrivilegeEscalation != false (container \"brig-index\" must set securityContext.allowPrivilegeEscalation=false), unrestricted capabilities (container \"brig-index\" must set securityContext.capabilities.drop=[\"ALL\"]), runAsNonRoot != true (pod or container \"brig-index\" must set securityContext.runAsNonRoot=true), seccompProfile (pod or container \"brig-index\" must set securityContext.seccompProfile.type to \"RuntimeDefault\" or \"Localhost\")"
  Error: failed post-install: 1 error occurred:
  	* job brig-index-migrate-data failed: BackoffLimitExceeded

This PR doesn't change helm charts or values at all, especially not the security context stuff. How can it be realted to this branch? Or is this a red herring?

[fisx]

The CI errors consistently happen on this branch and nowhere else, but I don't understand:

STDERR:
  I0309 10:14:26.035245    2534 warnings.go:110] "Warning: would violate PodSecurity \"restricted:latest\": allowPrivilegeEscalation != false (container \"brig-index\" must set securityContext.allowPrivilegeEscalation=false), unrestricted capabilities (container \"brig-index\" must set securityContext.capabilities.drop=[\"ALL\"]), runAsNonRoot != true (pod or container \"brig-index\" must set securityContext.runAsNonRoot=true), seccompProfile (pod or container \"brig-index\" must set securityContext.seccompProfile.type to \"RuntimeDefault\" or \"Localhost\")"
  Error: failed post-install: 1 error occurred:
  	* job brig-index-migrate-data failed: BackoffLimitExceeded

This PR doesn't change helm charts or values at all, especially not the security context stuff. How can it be realted to this branch? Or is this a red herring?

the logs contain network errors involving cassandra. but i'm only changing the migration logic in this PR, nothing in helm charts or values where PodSecurity would mean anything.

the riddle deepens...