Support for multi-cluster E2E Test framework

[sebastienvas]

Details on requirements are @ http://goo.gl/AQARBs#heading=h.wgf09th1ifw0
but basically:
The test framework should be able to:

1. Request 1 or more K8s clusters during test setup. Calls should block until clusters are available
2. Release all clusters requested by the test at test teardown
3. We should have a pool of clusters for integration testing

[ldemailly]

In 0.3 we're going to start supporting multi cluster and hybrid so we will need to test that we can join a mesh across clusters. At minimum that means some tests can request from test infra 2 clusters (clusterA and clusterB) and deploy on both and test reachability etc

@gnirodi can also add more

[danehans]

@ldemailly do you have a design doc, PRs, or any other pointer to learn more about 0.3 multi-cluster support?

[gnirodi]

http://goo.gl/AQARBs will be the canonical design doc for Istio multi-cluster support for v0.3
It's still WIP

[ldemailly]

Doc above, the short version is we will start with a pair of clusters (2) + 1 or 2 VMs
and that should be fine for e2e tests

[mattymo]

@ldemailly Can this doc made readable to the public?

[gnirodi]

Hi Matthew, The doc is publicly shared with the Istio community. I have additionally shared it with you via matthew.mosesohn@gmail.com. Please let me know if this does not work for you. You may want to join https://istio.io/community/, that way you get access to referenced docs as well :-) Thanks, Gautam

…

On Thu, Oct 19, 2017 at 6:05 AM, Matthew Mosesohn ***@***.***> wrote: @ldemailly <https://github.com/ldemailly> Can this doc made readable to the public? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#626 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AR45K4F5i5Z9nBRsypK4ZTLlWkUpZ22iks5st0kGgaJpZM4PIShn> .

-- Gautam Istio Heterogeneous Clouds

[danehans]

@gnirodi thank you for the pointer to the google design doc for multi-cluster. I do not see any PR references here or in the design doc for the work that needs to be accomplished. Are there any PR's in flight? If so, can I get a pointer? I would like to start contributing, but I want to avoid duplicating efforts. It may be helpful to create a label (i.e. /area multicluster) that can be used to track related issues/pr's associated to this effort. I look forward to contributing.

[gnirodi]

OK I've updated the design doc http://goo.gl/AQARBs#heading=h.wgf09th1ifw0
I'll assign this issue to myself until the design is complete.
Implementation may be farmed out to various folks and I'll move assignee accordingly.

[gnirodi]

The doc has a reasonably outlined design criteria.
yutongz has volunteered to help flesh out the detailed design around test infrastructure. So adding him to this issue.

[ldemailly]

note about the usage of "zone" in the Issue title, that terminology can be confusing:

we should test both
a) 1 kubernetes cluster spanning availability zones (datacenters) - a production readiness task
b) 2 kubernetes cluster (and 1 VM) in a single mesh

[sara4dev]

were you guys able to test the above #1 & #2? We have use case for both.

[yutongz]

/unassign @yutongz

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 30 days. It will be closed in the next 2 weeks unless it is tagged "help wanted" or other activity occurs. Thank you for your contributions.

[ldemailly]

Andy/costin update ?

[stale]

This issue has been automatically marked as stale because it has not had activity in the last 30 days. It will be closed in the next 2 weeks unless it is tagged "help wanted" or other activity occurs. Thank you for your contributions.

[stale]

This issue has been automatically closed because it has not had activity in the last month and a half. If this issue is still valid, please ping a maintainer and ask them to label it as "help wanted". Thank you for your contributions.