[Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Wang, Yi C yi.c.wang at intel.com
Wed Jul 24 02:07:32 UTC 2019


Hi Bin,

I can't ping my local registry too. I didn't check iptables rules of controller nodes. But I guess icmp was blocked by some rules. I can connect to external hosts by ssh from my worker nodes.

You can refer to below link on local registry deploy.
https://docs.docker.com/registry/deploying/

Thanks.
Yi
From: Yang, Bin [mailto:Bin.Yang at windriver.com]
Sent: Wednesday, July 24, 2019 9:47 AM
To: Wang, Yi C <yi.c.wang at intel.com>; Xie, Cindy <cindy.xie at intel.com>; starlingx-discuss at lists.starlingx.io
Subject: RE: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

And could you kindly educate me how to setup a local docker register over oam network? The network latency results a lot of issue during my installation of stx controller nodes.

Thanks

Best Regards,
Bin Yang,    Solution Engineering Team,    Wind River
ONAP Multi-VIM/Cloud PTL
Direct +86,10,84777126    Mobile +86,13811391682    Fax +86,10,64398189
Skype: yangbincs993

From: Yang, Bin
Sent: Wednesday, July 24, 2019 9:46 AM
To: 'Wang, Yi C'; Xie, Cindy; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Hi Yi,

Thanks for the response.

I did test if worker node could access to oam network by ping the docker proxy ip, and the result turned out that worker node cannot reach to that ip. I did the same test over controllers and the controllers could reach to that ip.

So I guess this is not an issue of docker settings, it might be something wrong with controllers if controllers are deemed to NAT the traffic from worker nodes to OAM network.

Any suggestion on how to check that?

Thanks

Best Regards,
Bin Yang,    Solution Engineering Team,    Wind River
ONAP Multi-VIM/Cloud PTL
Direct +86,10,84777126    Mobile +86,13811391682    Fax +86,10,64398189
Skype: yangbincs993

From: Wang, Yi C [mailto:yi.c.wang at intel.com]
Sent: Wednesday, July 24, 2019 9:26 AM
To: Yang, Bin; Xie, Cindy; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Hi Bin,

Worker nodes can access external network through controller nodes. In our deployment, we use a local registry which is accessible via OAM. Worker nodes do can pull images from the local registry. So I suggest you check

1. if your worker nodes can access external network through oam
2. check the docker configuration "/etc/docker/daemon.json" on your worker nodes.

Thanks.
Yi
From: Yang, Bin [mailto:Bin.Yang at windriver.com]
Sent: Wednesday, July 24, 2019 8:21 AM
To: Xie, Cindy <cindy.xie at intel.com<mailto:cindy.xie at intel.com>>; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: Re: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Hi Cindy,

Thanks for the response. I do locate in China, and I have setup a proxy to overcome the connectivity issue. That was verified to work by installing controllers, and by executing docker pull image on controller nodes.

The issue I am experiencing is that 'a worker node' have no access to oam network but trying to pull docker image while I am trying to deploy openstack helm using 'system application-apply stx-openstack' commands.

Thanks

Best Regards,
Bin Yang,    Solution Engineering Team,    Wind River
ONAP Multi-VIM/Cloud PTL
Direct +86,10,84777126    Mobile +86,13811391682    Fax +86,10,64398189
Skype: yangbincs993

From: Xie, Cindy [mailto:cindy.xie at intel.com]
Sent: Tuesday, July 23, 2019 10:00 PM
To: Yang, Bin; starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: RE: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Hi, Bin
I guess you're located in China and have firewall blocking some docker image, right? You may have to setup your local registry in your lab.

Thx. - cindy

From: Yang, Bin [mailto:Bin.Yang at windriver.com]
Sent: Tuesday, July 23, 2019 8:53 AM
To: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: [Starlingx-discuss] [Containers] Unable to pull docker image from a worker node

Dear experts,

I have been trying to install stx milestone3 over virtualbox with standard modes, I managed to install 2 controller nodes and 1 worker node, and provisioned them according to the instructions on wiki.  Then I uploaded openstack helm charts then apply it, then it failed to accomplish that operation.

As I investigate the root cause, I found out that the worker node is not in ready status:


[sysadmin at controller-0 ~(keystone_admin)]$ kubectl get  nodes
NAME           STATUS     ROLES    AGE     VERSION
compute-0      NotReady   <none>   3h32m   v1.13.5
controller-0   Ready      master   14d     v1.13.5
controller-1   Ready      master   13d     v1.13.5

The root cause is that worker node is lacking of access to external network directly, hence unable to pull docker image via the proxy:


[sysadmin at controller-0 ~(keystone_admin)]$ kubectl -n kube-system describe pods  kube-proxy-pftk8



Events:

  Type     Reason                  Age                    From                Message

  ----     ------                  ----                   ----                -------

  Warning  FailedCreatePodSandBox  38s (x462 over 3h36m)  kubelet, compute-0  Failed create pod sandbox: rpc error: code = Unknown desc = failed pulling image "k8s.gcr.io/pause:3.1": Error response from daemon: Get https://k8s.gcr.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)





compute-0:~# docker pull k8s.gcr.io/pause:3.1

Error response from daemon: Get https://k8s.gcr.io/v2/: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)

The traceroute to the docker proxy :

compute-0:~# cat /etc/systemd/system/docker.service.d/http-proxy.conf
[Service]
Environment="HTTP_PROXY=http://128.224.230.5:9090"
Environment="HTTPS_PROXY=http://128.224.230.5:9090"
Environment="NO_PROXY=localhost,127.0.0.1,registry.local,192.168.204.2,192.168.204.3,10.0.2.25,10.0.2.26,192.168.204.4,10.0.2.27"


compute-0:~# traceroute 128.224.230.5
traceroute to 128.224.230.5 (128.224.230.5), 30 hops max, 60 byte packets
1  controller-0 (192.168.204.3)  0.200 ms  0.256 ms  0.208 ms
2  * * *
3  * * *

Can anybody help explain how a worker node without access to external network could pull docker images? How can I workaround this issue?

Thanks

Best Regards,
Bin Yang,    Solution Engineering Team,    Wind River
ONAP Multi-VIM/Cloud PTL
Direct +86,10,84777126    Mobile +86,13811391682    Fax +86,10,64398189
Skype: yangbincs993

From: Perez Ibarra, Maria G [mailto:maria.g.perez.ibarra at intel.com]
Sent: Tuesday, July 23, 2019 7:49 AM
To: starlingx-discuss at lists.starlingx.io<mailto:starlingx-discuss at lists.starlingx.io>
Subject: [Starlingx-discuss] [Containers] Sanity Test - ISO 20190722

Status of the Sanity Test for last CENGN ISO: bootimage.iso from 2019-Jul-22 (link<http://mirror.starlingx.cengn.ca/mirror/starlingx/master/centos/20190721T233000Z/outputs/iso/>)



Status: RED



===========================================



Sanity Test is executed in a Containers - Bare Metal Environment


AIO - Simplex



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  49 TCs [PASS]

Sanity Platform   07 TCs [PASS]



TOTAL: [ 61 TCs PASS ]


AIO - Duplex



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  52 TCs [PASS]

Sanity Platform   07 TCs [PASS]



TOTAL: [ 64 TCs PASS ]



Standard - Dedicated Storage (2+2+2)



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  52 TCs [PASS] 24 TCs FAIL

Sanity Platform   09 TCs [PASS]



TOTAL: [ 66 TCs PASS ]


===========================================



Sanity Test is executed in a Containers - Virtual Environment


AIO - Simplex



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  49 TCs [PASS]

Sanity Platform   07 TCs [PASS]



TOTAL: [ 61 TCs PASS ]



AIO - Duplex



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  52 TCs [PASS] 24 TCs FAIL

Sanity Platform   07 TCs [PASS]



TOTAL: [ 64 TCs PASS ]



Standard - Local Storage (2+2)



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  52 TCs [PASS]

Sanity Platform   08 TCs [PASS]



TOTAL: [ 65 TCs PASS ]



Standard - External Storage (2+2+2)



Setup             04 TCs [PASS]

Provisioning      01 TCs [PASS]

Sanity OpenStack  52 TCs [PASS] 24 TCs FAIL

Sanity Platform   08 TCs [PASS]



TOTAL: [ 65 TCs PASS ]



-----------------------------------------------------------------------------------



Create instance from Image or from Volume fails https://bugs.launchpad.net/starlingx/+bug/1837241

Regards
Maria G.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.starlingx.io/pipermail/starlingx-discuss/attachments/20190724/b5c94947/attachment-0001.html>


More information about the Starlingx-discuss mailing list