Router sharding proposal #506
Conversation
|
@smarterclayton your comments on https://github.com/pweil-/origin/pull/1 have been addressed here. |
|
@smarterclayton @pmorie I will be closing https://github.com/pweil-/origin/pull/1 so there is no confusion. |
| type Route { | ||
| // other fields not shown | ||
| RouterDNS string | ||
| AllocationStatus RouteAllocationStatus |
There was a problem hiding this comment.
This should follow the upstream convention:
Status RouteStatus
type RouteStatus struct {
DNS string
}
Fields that are set after creation are not part of "spec".
|
@smarterclayton your comments on allocation and scheduling should be addressed now. |
|
@smarterclayton @pweil-'s changes for DNS are merged into this PR |
|
@abhgupta @rajatchopra @akram plz review |
|
@pweil- we should add content dealing with the specifics of which actors do things on behalf of a route to set up DNS |
|
@pmorie At this point it is intentionally ambiguous and notes that a plugin or manual setup could be used to set up DNS on behalf of a routes. Since the use case will not be covering custom DNS records at this point (setup is left to the user) we will only be configuring wild cards for shards. There is an example in the file but we're expecting administrators to know how to configure their existing DNS infrastructure. If we decide to make a plugin the document will need to be updated with more specifics just like the router portion. In fact I'd venture that the design will be exactly like the router portion |
pmorie
force-pushed
the
router-shards-proposal
branch
from
074fd25 to
5e906d1
Compare
|
@danmcp care to tag this? |
|
[merge] |
|
Origin Merge Results: SUCCESS (https://ci.openshift.redhat.com/jenkins/job/merge_pull_requests_openshift3/507/) (Image: devenv-fedora_466) |
|
Evaluated for origin up to 5e906d1 |
…uder Do not exclude the excluder for atomic-openshift
No description provided.