d63c47149d
integration: Fix broken test in externallb
...
Running `terraform validate` on `externallb/kubernetes.tf` was failing
because resource name can't contain ":".
2018-10-20 18:27:41 +03:00
7dca8a78d2
terraform: Fix resource formatting for IPv6 CIDRs
...
Solved #5976
2018-10-20 15:58:42 +03:00
e119068480
Update tests for channel bump
2018-10-15 16:56:34 -04:00
646fee8a9d
Merge pull request #5857 from justinsb/strict_cloudformation_output_checking
...
Stop cloudformation output switching to literal quotes
2018-10-02 22:53:19 -07:00
3fe0287ff8
Merge pull request #5862 from justinsb/follow_on_5744
...
Follow on for #5744
2018-10-02 17:22:03 -07:00
b3d61542cd
Merge pull request #5820 from justinsb/etcd_manager_channels_step1
...
etcd: introduce field to specify whether we are using etcd-manager or legacy mode
2018-10-02 14:20:37 -07:00
1e2a62992b
Use JoinSuffixes for node->master traffic, also fix AmazonVPC rule
...
This ensures we are consistently naming our rules
2018-10-02 11:53:41 -07:00
1906bcdf5d
We need to create the cross-product of rules for SG overrides
...
e.g. each master SGs need to be configured to talk to each master SG
2018-10-02 11:53:41 -07:00
1f2a8042b5
Test case where we mix override & default SGs
2018-10-02 11:53:41 -07:00
bfb54935ff
Build security groups along with suffixes
...
Fixes the case where we mix use of specified & default SGs.
2018-10-02 11:53:41 -07:00
076742f528
Still materialize terraform output in tests
...
If we skip it, we can't test it. We do expect that most users will
use a lifecycle that only warns though.
2018-10-02 11:53:39 -07:00
4cd8dbae3f
Update tests
2018-10-02 01:42:46 -07:00
a82f548ff8
Allow using existing/shared Security Groups
...
Verbosely log when a user overwrites LB or IG security groups
Change SecurityGroup to SecurityGroupOverride
Allow using existing/shared Security Groups
Update tests
2018-10-02 00:51:39 -07:00
1ad7a84d53
Stop cloudformation output switching to literal quotes
...
yaml can't store the \r character, so we strip the \r characters in
the expected output to keep the yaml output in block-quote style.
Also don't Fatalf out of an error, rather Errorf so we print all
problems.
2018-10-01 20:44:19 -07:00
54c499fe9b
Introduce field to specify whether we are using etcd-manager or legacy mode
...
Splitting this out from the bigger bundle PR
2018-10-01 15:36:24 -07:00
2fc634c556
Revert "Apply cloud labels into ELB"
...
This reverts commit b52c945f9f
2018-09-25 13:53:38 +01:00
e6bf6b9f32
Field names are case-sensitive again
...
There was a regression in apimachinery which meant that kubernetes
tolerated field names with incorrect case. Upstream bug is
https://github.com/kubernetes/kubernetes/issues/64612
Syncing up with latest kubernetes will mean we get the same breaking
change as kubernetes has/had. It should only affect people that are
manually building YAML / JSON.
Added as a significant item to release notes.
2018-09-23 19:23:28 -04:00
237043dded
Update expected test output for script changes
2018-09-21 14:51:45 -04:00
b1c446f8f3
Merge pull request #5503 from mikesplain/fix_suspendprocess
...
Fix suspendprocess
2018-09-03 16:28:31 -07:00
2319db8f59
Merge pull request #4797 from justinsb/check_no_legacy_tags_on_shared
...
lifecycle tests: check no legacy tags on shared resources
2018-09-03 15:28:26 -07:00
c61fb537ac
Merge pull request #5593 from wingyplus/elb_cloud_labels
...
Apply cloud labels into ELB
2018-08-18 02:00:01 -07:00
16985c3abc
Remove _kubernetes_master tag
...
We can get the master role just as readily from the InstanceGroup spec
2018-08-14 21:01:07 -04:00
1c3949bdfe
Don't assume that we only have one subnet per AZ
...
I made a mistaken assumption in
dde2100a19#5587 this was not the case.
What I was trying to achieve was not to include the cluster name, so
for the case of subnets this commit just uses the subnet name from the
cluster spec, which should be unique and stable. That is hopefully at
least as meaningful.
Thankfully we hadn't released a version with the erroneous naming.
Fix #5587
2018-08-13 13:15:03 -04:00
b52c945f9f
Apply cloud labels into ELB
...
Fixes #2048
2018-08-11 00:43:19 +07:00
a7b22b4876
Remove GetAsgForInstance IAM permission
...
It isn't a valid IAM permission - it was introduced in error, but IAM
is kind enough to ignore it.
Fixes #5549
2018-08-02 11:27:29 -04:00
38fd71605e
Add test for additionalPolicies
2018-07-27 00:18:53 -04:00
a93ca798ec
Fix tests
2018-07-23 20:55:44 -04:00
00cd85d73e
Add autoscaling group ids to terraform module output
2018-07-20 13:15:03 +02:00
fd70778ccf
Fix tests that crossed during PR merges
2018-07-19 22:07:02 -04:00
630ea429ae
Merge pull request #4677 from usabilla/external-load-balancers
...
Add the ability to specify external loadbalancers for instancegroups
2018-07-19 21:54:00 -04:00
c5c2bd1acb
Don't repeatedly download nodeup
...
Only delete if the file doesn't match the hash.
Should help with retry issues / bandwidth
2018-07-19 12:09:13 -04:00
978f66266c
Generate locals for terraform target
2018-07-13 22:28:27 +02:00
8256c3cc73
Fixed integration tests after rebasing
2018-06-29 12:42:06 +02:00
2c5edfe992
Added integration test for external load balancers
2018-06-29 11:39:21 +02:00
b066b5b8d6
Merge pull request #5359 from justinsb/revert_cos_channel
...
Revert COS in stable/alpha channel
2018-06-20 20:54:07 -07:00
587d8374b5
Revert COS in stable/alpha channel
...
No problem with COS per-se, but these versions have the newer docker,
which includes the --storage flag. We fixed that in master in #5258 ,
but older versions of kops - including the currently released version
1.9.1 - don't have the fix.
Revert to fix the problem immediately, but opened #5358 to track a
more realistic fix.
2018-06-20 20:00:15 -04:00
1af610a6c4
Use fast-retry and cleanup task options for tests
...
The launch configuration test exposed that our integration tests don't
retry for very long, and wait a long time in between retries.
Create a RunTasksOptions type to hold the parameters, in particular
max task time, and the amount of time we wait when all tasks have
failed.
2018-06-18 00:04:23 -04:00
e2debd99a6
Update expected TF output for latest master
2018-06-11 11:12:23 -04:00
93b5e24801
Merge pull request #5030 from chrissnell/export-addl-tf-outputs
...
Export outputs to aid with VPC peering in Terraform
2018-06-11 11:11:52 -04:00
1b29b54d5b
Merge pull request #4171 from rifelpet/rifelpet_iam
...
Add support for external IAM Instance Profiles
2018-06-10 14:54:00 -07:00
a57a6de4a8
Merge pull request #5139 from justinsb/fix_subnet_egress_indexing
...
Fix issue where we assumed that private zone were in order
2018-06-09 09:27:01 -07:00
5f0b63100d
Add support for using existing instance profiles
2018-06-08 10:33:09 -07:00
78b81dd172
Merge branch 'master' into export-addl-tf-outputs
2018-06-08 12:22:43 -05:00
ef5936d8b5
Support overlay2 in docker
...
We also have to stop passing the flag on ContainerOS, because it's set
in /etc/docker/default.json and it's now an error to pass the flag.
That in turn means we move those options to code, which are the last of
those legacy config options. (We still have a few tasks declaratively
defined though)
2018-06-07 17:11:11 -04:00
4a131f2a50
Implement AdditionalCIDR configuration on non-shared VPCs
2018-06-05 01:12:06 -07:00
e24aa79aa9
Updated terraform expected test output
2018-06-04 10:52:08 -07:00
6741158c97
Merge pull request #5056 from povilasv/gce-ssh
...
Add public ssh keys for GCE
2018-06-03 10:59:53 -07:00
e73ee4e3e8
Update tests for alpha -> stable promotion
2018-05-31 07:41:31 -07:00
bc79f0a69b
tests: failing test when multiple egress
...
Also requires that the subnets be out of order, which is likely why we
didn't hit it in practice.
2018-05-10 13:28:05 -04:00
02ab1f90be
Merge pull request #5133 from relu/support-ec2-instance-nat-egresses
...
Add ability to use ec2 nat instance as egress
2018-05-10 08:22:53 -07:00
Ariel Mashraki