But this place has already been changed in my patch

https://review.opendev.org/714894

Zhipeng

From: Richard, Joseph <Joseph.Richard@windriver.com>
Sent: 2020年3月25日 22:08
To: Liu, ZhipengS <zhipengs.liu@intel.com>
Cc: starlingx-discuss@lists.starlingx.io
Subject: Re: [Starlingx-discuss] IPv6 issue when apply OpenStack on IPv6 simplex.

I see at least one place [1] where keystone is configured to listen on 0.0.0.0, so that's probably why.

[1] https://github.com/openstack/openstack-helm/blob/master/keystone/values.yaml#L949

---

From: Liu, ZhipengS <zhipengs.liu@intel.com>
Sent: Wednesday, March 25, 2020 5:19 AM
To: Richard, Joseph <Joseph.Richard@windriver.com>
Cc: starlingx-discuss@lists.starlingx.io <starlingx-discuss@lists.starlingx.io>
Subject: RE: [Starlingx-discuss] IPv6 issue when apply OpenStack on IPv6 simplex.

Hi Josesh,

I have added 2 more patches for your review.

https://review.opendev.org/714894

https://review.opendev.org/714898

So far, it can deploy services until heat. But I still have an intermittent issue that

***-ks-user jobs always need restart several times to get task finished.

Sometimes, it may cause endpoint job failure and block deployment procedure.

Log looks like below, still not find the root cause. Could you help on this point? Thanks!

Failed to discover available identity versions when contacting http://keystone-api.openstack.svc.cluster.local:5000/v3. Attempting to parse version from URL.

Unable to establish connection to http://keystone-api.openstack.svc.cluster.local:5000/v3/auth/tokens: HTTPConnectionPool(host='keystone-api.openstack.svc.cluster.local', port=5000): Max retries exceeded with url: /v3/auth/tokens (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f857a551250>: Failed to establish a new connection: [Errno 101] Network is unreachable',))

Thanks!

Zhipeng

From: Liu, ZhipengS
Sent: 2020年3月23日 21:50
To: 'Richard, Joseph' <Joseph.Richard@windriver.com>
Cc: 'starlingx-discuss@lists.starlingx.io' <starlingx-discuss@lists.starlingx.io>
Subject: RE: [Starlingx-discuss] IPv6 issue when apply OpenStack on IPv6 simplex.

Hi Joseph,

I have updated my 2 patches, both tested on ipv4 and ipv6 setup.

https://review.opendev.org/#/c/714034/

https://review.opendev.org/#/c/714038/

But still failed when deploy keystone, glance, cinder later.

Common failure causes are found that need to bind host to :: instead of 0.0.0.0.

Otherwise, it will only bind port to ipv4 address, which causes db connect failure or pod probe failure.

After below changes, it can run until deploying cinder.

1)Change mariadb   value.yaml    

conf:  bind adress  from  0.0.0.0 to ::    

2) Change keystone   value.yaml  

  wsgi_keystone: |

    {{- $portInt := tuple "identity" "internal" "api" $ | include "helm-toolkit.endpoints.endpoint_port_lookup" }}

     Listen 0.0.0.0:{{ $portInt }}   change to Listen :::{{ $portInt }}

3) Change glance value.yaml

    Conf.Default:

    add bind_host = "::"

I have to check these services deploy one by one.

Thanks!

Zhipeng