Choosing a port in advance is racy. A better solution is to use a Unix Domain
socket in the per-etcd-instance data directory. Then the name can be determined
in advance and there's no risk of conflicts with other etcd instances.
With unix:// for the endpoint, we have to be a bit more careful about
passing a TLS config to the etcd client library because for unix://, in
contrast to http://, it tries to use an incomplete config which
then fails to establish the connection.
The new etcd balancer (>3.3.14, 3.4.0) uses an asynchronous resolver for
endpoints. Without "WithBlock", the client may return before the
connection is up.
Signed-off-by: Gyuho Lee <leegyuho@amazon.com>
This change updates the ETCD storage test so that its data is
exported. Thus it can be used by other tests. The dry run test was
updated to consume this data instead of having a duplicate copy.
The code to start a master that can be used for "one of every
resource" style tests was also factored out. It is reused in the
dry run test as well.
This prevents these tests from drifting in the future and reduces
the long term maintenance burden.
Signed-off-by: Monis Khan <mkhan@redhat.com>
Many integration tests delete all keys in etcd as part of their cleanup.
To run these tests in parallel we must run several etcd daemons, each on
different port and pass etcd url to the test suite.
* Add an allocator which saves state in etcd
* Perform PortalIP allocation check on startup and periodically afterwards
Also expose methods in master for downstream components to handle IP allocation
/ master registration themselves.
