b0c5002552
tls_test: include client TLS
2017-05-17 13:36:19 -07:00
d47bd08edd
upgrade_test: re-enable per cluster S3 tests
2017-05-17 11:08:35 -07:00
45f4e360af
upgrade_test: disable per cluster S3 policy for 0.2.x
...
This is new feature we introduced in current master/0.3.x .
Gonna enable test again from 0.3.x on.
2017-05-13 17:11:20 -07:00
8d01f4d9cb
*: per cluster s3 policy
2017-05-11 16:11:44 -07:00
6c2d0e4277
upgradetest: remove other backupinterval
2017-05-11 16:10:57 -07:00
c45b4f8da0
upgradetest: remove backup interval
2017-05-11 14:56:19 -07:00
b1726f3602
upgradetest: Add disaster recovery test
2017-05-10 21:15:40 -07:00
b56abe3b1b
Merge pull request #1080 from hasbro17/haseeb/op-upgrade-restore
...
upgradetest: Add test to restore from backup of old cluster
2017-05-09 15:17:11 -07:00
a3506103c4
upgradetest: Add test to restore from backup of old cluster
2017-05-09 14:07:11 -07:00
84ee7b587c
cluster: dynamically update backup policy
2017-05-09 13:37:21 -07:00
3bb48d12a2
e2e: separate cluster and backup deletion checking
2017-05-08 10:26:22 -07:00
0868999a39
upgrade_test: add TestBackupForOldCluster
2017-05-07 21:38:38 -07:00
2c63aa5c57
e2eutil: add WaitPodsDeleted util families
2017-05-06 23:34:11 -07:00
07393f6ba5
Merge pull request #1063 from hasbro17/haseeb/op-upgrade-heal-1-member
...
e2e/upgradetest: 1 member recovery test
2017-05-06 14:06:22 -07:00
c49ec00216
e2e/upgradetest: 1 member recovery test
2017-05-06 12:28:42 -07:00
11324625ed
e2eutil: port WaitBackupPod() and MakeBackup()
2017-05-05 12:23:12 -07:00
caf1efafef
e2eutil: port killMembers
2017-05-05 11:18:16 -07:00
4d77895879
upgrade_test: fix delete operator
...
1. should defer delete operator after create
2. reduce grace period
2017-05-05 10:21:09 -07:00
c130e6afa0
Merge pull request #1043 from hasbro17/haseeb/op-upgrade-resize-add-wait
...
upgradetest/TestResize: wait for cluster TPR, remove unnecessary waits
2017-05-04 14:29:17 -07:00
c79727aa93
s3: refactor New() logic and hide internal dir layout
...
We have expose dir layout in e2e test code.
We need to refactor existing translate layer code and hide it.
2017-05-04 11:29:25 -07:00
6496ac3501
upgradetest/TestResize: wait for cluster TPR, remove unnecessary waits
2017-05-04 10:18:40 -07:00
21befb07a3
e2e: test should always pull latest image
...
We might not use versioned image in test, e.g. nightly build dev image.
2017-05-03 14:21:03 -07:00
5265850787
e2e: fix wait s3 delete check path
2017-05-03 13:24:43 -07:00
235640b724
e2e/util.go: Move waitUntilSize functions to e2eutil for common use
2017-05-02 15:37:55 -07:00
3f9834ea99
upgrade_test: make gen-name have "-" at the end as prefix
2017-05-02 14:34:02 -07:00
6261483235
e2e/util.go: Move etcd spec related methods into e2eutil for common use
2017-05-02 13:47:16 -07:00
17a856693d
e2e/util.go: refactor DeleteCluster() for common use between different frameworks
2017-05-02 12:34:27 -07:00
9d9301565a
backup: use Deployment to manage sidecar instead od ReplicaSet
...
There are also code that handles backward compatibility:
```
Once operator is upgraded, all backup-enabled cluster will go through
an upgrade process that deletes backup sidecar's ReplicaSet and creates
new Deployment for sidecar. If upgrading failed for any reason, cluster
TPR's `status.phase` will be FAILED. Recreate of the cluster TPR is
needed on failure case.
```
2017-05-02 10:57:41 -07:00
6ecbe2d901
e2e/util.go: refactor CreateCluster() into e2e/e2eutil for common use
2017-05-01 14:28:12 -07:00
87947ceaa3
tpr_util,e2e/util: AtomicUpdateClusterTPR to retry safe update
2017-05-01 11:56:45 -07:00
d803a15a5f
upgradetest: fix compile
2017-04-29 15:00:41 -07:00
3da1fc831a
Merge pull request #1023 from hasbro17/haseeb/remove-unconditional-updateTPR
...
e2e: safe update TPR instead of unconditional update
2017-04-28 15:28:37 -07:00
c56fe48623
e2e: safe update TPR instead of unconditional update
2017-04-28 15:01:53 -07:00
510adfb7cf
hack/test: add upgrade test
2017-04-28 13:46:36 -07:00
4d22063599
upgradetest: add resize test
...
Done the basic workflow of an upgrade test:
- create operator
- create cluster
- upgrade operator
- resize the cluster works
The code isn’t modular yet but we will improve it in next PR.
Gonna add a script and set up a job to run upgrade test too.
2017-04-28 12:52:50 -07:00
0bc16aa6cf
upgradetest: add framework for operator deploy
2017-04-28 10:25:53 -07:00
63696746ee
add unused and gosimple checks
2017-04-24 16:53:48 -07:00
7f12589890
Merge pull request #980 from hongchaodeng/f2
...
e2e: work around k8s 1.6 grace period issue
2017-04-21 14:54:10 -07:00
c1566b82b9
*: get rid of /bin/sh in container commands
2017-04-21 12:50:32 -07:00
a039719057
e2e: kill the last pod in recovery test
...
The last pod could have not come up yet.
A simple solution is to kill the last pod.
2017-04-21 10:33:59 -07:00
4b45db451e
e2e: work around k8s 1.6 grace period issue
2017-04-21 10:24:17 -07:00
83250bbc10
*: update client-go to 3.0-beta
2017-04-19 22:42:33 -07:00
27e457e61a
TLS: update docs (client communication)
...
Update the user docs to reflect operator's client key/cert and
etcd server's client key/cert.
Meanwhile, change the staticTLS fields to be pointer so that we preserve
the possibility of more options.
2017-04-19 07:28:01 +08:00
6698f7adeb
e2e: add tls test
2017-04-15 08:07:27 +08:00
e1d3ca540b
e2e: group status test and speed up with parallel
2017-04-07 07:14:51 +08:00
c5ce5e2e18
contorller: add test to cover events for failed cluster
2017-03-31 22:18:52 +08:00
659d8d80ff
*: update default etcd version to 3.1.4
...
There is a hang process being fixed:
https://github.com/coreos/etcd/issues/7512
2017-03-22 10:45:16 -07:00
753bbe7a76
*: add errexit to all sh command
2017-03-13 09:04:00 -07:00
3e127e4c99
group ReadyMembers/UnreadyMembers together
2017-03-09 15:33:42 -08:00
1c05923a46
Merge pull request #874 from hongchaodeng/v
...
SelfHosted: use hostPath for etcd pod
2017-03-09 14:30:54 -08:00
Hongchao Deng
hasbro17
Hongchao Deng
Xiang