Closed scsmithr closed 1 year ago
Would require some coordination on cloud side of things to ensure connectivity. Maybe try reaching /healthz
?
/healthz
seems absolutely reasonable, but given that this is sending to internal
, it can honestly be anything there as well.
If it simply just needs evidence of a ping/pong, healthz
is the way to go
I'm adding a "Cloud Client" in https://github.com/GlareDB/glaredb/pull/324 which may be useful here.
I'm adding a "Cloud Client" in #324 which may be useful here.
Going to remove my assignee until after that point. I should really only pick up things in GlareDB/glaredb
that are either at my level or that it makes sense to, at this point.
It seems like something that should proceed this is a refactor of pgsrv
to use the cloud
client. Does that sound right?
Yep, we should use CloudClient
for this. Just switching over to CloudClient
will close this out since it does a ping on create.
Do #367 first.
Stale - can reopen later
Have a check that checks pgsrv can hit cloud.
It may be easiest to start something on a dedicated port to prevent needing to shoehorn a check into our existing protocol code (which we could do with an extended protocol).