Github user elilevine commented on the pull request:
https://github.com/apache/phoenix/pull/104#issuecomment-129690118
Thanks for the details, @JeffreyLyonsD2L. I have to say I'm still torn on
whether or not this Jira makes much sense. What if I told you that your concern
#2 is already addressed? Currently tenant id columns can be either VARCHAR or
fixed-length CHAR. Does that make it easier for you to use multi-tenancy in
Phoenix?
Regarding your points #1 and 3, you are forced to do the conversion once
per Connection when you set the TenantId parameter either way. Not sure what
kind of conversion you mean beyond that. Normally when you read/write via a
tenant-specific connection you only set TenantId on the Connection and don't
need to actually write it or read it. Are you working with this data over
non-tenant-specific connections?
---
I went though your code again. Everything looks good. The only thing
missing now is:
1. Tests that create tenant views with non-string tenant id columns and
test data upsertion/retrieval.
2. Same as above but negative. Test what happens when conversion fails.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
