URL:
https://github.com/SSSD/sssd/pull/136
Title: #136: Tlog integration
pbrezina commented:
"""
So the question is why we first search the cache then do the request?
The current code is to notify nss responder about changes:
1. Call `sysdb_initgroups` and remember the groups that are cached at the moment.
2. Update the cache with the content from LDAP.
3. Send original groups to NSS responder so it can decide if anything has changed. If so,
it invalidates in-memory cache for initgroups.
Thinking about it, this is something that can be moved to NSS responder now. Anyway...
What you want to achieve is:
1. Update the cache
2. Call `sysdb_initgroups` to fetch the groups
3. Match the groups.
I didn't realize it during the review that you are not doing it in this way, sorry
about that. So you don't have really that much in common with the NSS request, you
just want to hook into the postprocess function which is called when the request is
finished, before the result is send to the responder.
Does this answer your question?
"""
See the full comment at
https://github.com/SSSD/sssd/pull/136#issuecomment-294822182