New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix containerized 3.9 -> 3.10 upgrade #8239
Conversation
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: vrutkovs The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
roles/etcd/defaults/main.yaml
Outdated
| @@ -89,7 +91,7 @@ etcd_listen_client_urls: "{{ etcd_url_scheme }}://{{ etcd_ip }}:{{ etcd_client_p | |||
| #etcd_peer: 127.0.0.1 | |||
| etcdctlv2: "{{ r_etcd_common_etcdctl_command }} --cert-file {{ etcd_peer_cert_file }} --key-file {{ etcd_peer_key_file }} --ca-file {{ etcd_peer_ca_file }} -C https://{{ etcd_peer }}:{{ etcd_client_port }}" | |||
|
|
|||
| etcd_service: etcd | |||
| etcd_service: "{{ (r_etcd_common_etcd_runtime == 'docker') | ternary('etcd_container', 'etcd') }}" | |||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think we need this. We're brute-force removing the services in favor of pods, except for stand-alone rpm etcd.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yep, this currently breaks, so I'll remove etcd commits
| @@ -13,5 +14,14 @@ | |||
| - reload systemd units | |||
| - restart node | |||
|
|
|||
| - block: | |||
| - name: Remove existing systemd service | |||
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure if this is what we want?
If the node is already a system container (which was possible previously in all types of deploys), and we remove this file and the host was already updated (IE, this is a no-op because this host was already upgraded) I'm not sure that the service unit will be placed again via the atomic command.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If the node is already a system container (which was possible previously in all types of deploys)
On Atomic installs this was a different node service. As a result atomic install would crash and complain that service file already exists. Atomic install is not very smart though - it won't create a service file if its missing during a second run.
This part is WIP still
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Reworked this to remove the services only when node is not yet bootstrapped
400a23d
to
7daae2a
Compare
e5ee8b8
to
7010f57
Compare
4b08c5c
to
b3dd1f8
Compare
b3dd1f8
to
713b401
Compare
54087ea
to
9a893cf
Compare
9a893cf
to
c5666cd
Compare
|
@vrutkovs: The following test failed, say
Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
373a881
to
7019c2c
Compare
9c6c5b3
to
bc70748
Compare
* Fix etcd runtime detection when setting up etcd * During upgrade etcd runtime should be detected using systemd services status * Mask, disable and stop services before removing service files * System container should be removed as stopping service doesn't seem to cut it * etcd cluster health check would wait for etcd static pod container to start
bc70748
to
f202390
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Also, I'd discussed w/ Clayton copying all the variables from /etc/sysconfig/{{ openshift_service_type }}-api to /etc/origin/master/master.env because at least in free-int we lost AWS configuration because that was added manually to /etc/sysconfig/atomic-openshift-master-api.
What do you think?
| state: absent | ||
| with_items: | ||
| - /etc/sysconfig/origin-master-api | ||
| - /etc/sysconfig/origin-master-controllers |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
- /etc/sysconfig/{{ openshift_service_type }}-api
- /etc/sysconfig/{{ openshift_service_type }}-controllers
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Fixed
|
BTW, migrating anything from /etc/sysconfig/atomic-openshift-master-* to master.env we can look at in a follow up. |
This seems to be set in current master.env. Lets fix the missing vars later on, as I'm not sure which ones need to be migrated and which are not required anymore (e.g. |
f202390
to
7dc9e74
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/lgtm
all concerns seem to have been addressed
|
bot, retest this please |
1 similar comment
|
bot, retest this please |
|
We need to dig into the flake we're seeing here, discussion on that at #8264 |
Fixes https://bugzilla.redhat.com/show_bug.cgi?id=1571724
TODO: