Rewrite ASM module #1140
Rewrite ASM module #1140
Conversation
|
Thanks for the PR! 🚀 |
Monkeyanator
force-pushed
the
rewrite-asm-module
branch
from
e910b17 to
a4d7560
Compare
Monkeyanator
force-pushed
the
rewrite-asm-module
branch
from
a4d7560 to
f5b65bd
Compare
Monkeyanator
requested review from
a team,
Jberlinsky and
bharathkkb
as code owners
Monkeyanator
force-pushed
the
rewrite-asm-module
branch
from
995b927 to
7ad8320
Compare
|
Addressed comments. CI failing on an issue that doesn't seem related to these changes though, any thoughts here? @bharathkkb |
|
@Monkeyanator I noticed it when working on another PR. You can switch to using the registry source for |
|
Opened #1159 since it was affecting a few PRs |
|
Thanks for looking into it @bharathkkb, made the change from #1159 here but seems like we still run into issues about resources changing (hard to pin down where the step actually fails from the logs). Think the PR with just the fix is failing CI as well |
|
@Monkeyanator seemed like the failure in #1159 was a flake. I have kicked it off again. Also could we add some kind of upgrade guide for users coming from the old module? Is there a possible upgrade path minimizing disruption? |
|
@bharathkkb the exposed options and implementation are different enough from the previous implementation where it'll be tough to support a migration path here. Is it possible to document to pin to the old version for this first iteration? Updated the README documenting this. |
Makes sense, I dropped a comment above to move that section to the upgrade guide. |
|
Super excited that this PR has been merged. Do you think it will solve this issue? #1114 |
Yes, the segfault probably comes from somewhere in the |
|
Why was asm_version removed? |
* Remove previous ASM module and add initial implementation of new * Move from kubernetes_manifest to bash CPR creation * Add meshconfig.googleapis.com service * Add namespace to CPR creation script wait * Change example to use google_container_cluster * Improve CPR status wait duration * Use retries rather than sleeping to wait for CPR CRD existence * Enable servicemesh feature in module * Revert changes to examples * Fix enable_mdp to just enable CNI * Lint * minor fixes * Bump timeout on status wait * Don't create MeshConfig if unset * Fix ASM sample * Update README autogen * Minor fixes * Remove meshConfig from module * fix end to end tests * lint fixes * Remove enable_mdp, add enable_vpc_sc and fleet_id * move VPC-SC from labels to annotations * update README * use default node pool size * use wip for exmaple cluster creation * add feature enablement back to module * lint * remove feature enablement from module * remove depends_on * fix unspecified channel bug * minor fixes * add more testing * update docs * change required module version * fix cclb * change registry source to run in CI * update CROSS_CLUSTER_SERVICE_DISCOVERY to multicluster_mode * update README * fix test * fix wording * remove from README * iterate on comments Co-authored-by: kaariger <[email protected]>
The existing ASM module is a near-1:1 wrapper around
install_asmwhich is not ideal. This rewritten module uses the ControlPlaneRevision KRM API to provision the control plane. We still break out and use a provisioner to apply the CPR (due to limitations around thekubernetes_manifestresource) but it's much slimmer than before, and doesn't involve curling down an installation script.