<div dir="ltr">Hi<br><br><div class="gmail_quote"><div dir="ltr">On Mon, Aug 13, 2018 at 8:40 PM Jason Montleon <<a href="mailto:jmontleo@redhat.com">jmontleo@redhat.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Someone recently filed a bug for the v3.10/latest version mismatch. We <br>
have some ideas how to handle it better and are working on implementing <br>
it for 3.11 (and possibly 3.10).<br>
<br>
For now it's pretty easy to modify the dc image to get the release you want.<br>
<br>
oc patch dc -n openshift-automation-service-broker <br>
openshift-automation-service-broker <br>
--patch='{"spec":{"template":{"spec":{"containers":[{"name": <br>
"penshift-automation-service-broker", <br>
"image":"ansibleplaybookbundle/origin-ansible-service-broker:release-1.2"}]}}}}'<br></blockquote><div><br></div><div>I tried applying the patch after my Ansible Service Broker installation was failing in the same exact way that Charles mentions, but the odd thing is that I don't have the <i>openshift-automation-service-b</i><i>roker namespace</i>. Is that to be expected or perhaps some bug?<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
release-1.0 corresponds to 3.7<br>
release-1.1 3.9<br>
release-1.2 3.10<br>
latest what will become release-1.3 3.11<br>
<br>
I'm not sure anything earlier than release-1.2 will work with the <br>
resources created by the APB so I'd stick to release-1.2 and latest if <br>
you can for now.<br>
<br>
On 08/13/2018 12:52 PM, Charles Moulliard wrote:<br>
> Is there a reason why we don't install the APB docker image tagged for <br>
> openshift 3.10 but instead latest  ?<br>
> <br>
> <a href="https://github.com/openshift/openshift-ansible/blob/release-3.10/roles/ansible_service_broker/defaults/main.yml#L26" rel="noreferrer" target="_blank">https://github.com/openshift/openshift-ansible/blob/release-3.10/roles/ansible_service_broker/defaults/main.yml#L26</a><br>
> <br>
> <br>
> <br>
> On Mon, Aug 13, 2018 at 6:42 PM, Charles Moulliard <<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a> <br>
> <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>>> wrote:<br>
> <br>
>     FYI, events still report that the pod is unhealthy<br>
> <br>
>     6:39:49 PMWarningUnhealthy Readiness probe failed: Get<br>
>     <a href="https://10.128.0.16:1338/healthz" rel="noreferrer" target="_blank">https://10.128.0.16:1338/healthz</a> <<a href="https://10.128.0.16:1338/healthz" rel="noreferrer" target="_blank">https://10.128.0.16:1338/healthz</a>>:<br>
>     dial tcp <a href="http://10.128.0.16:1338" rel="noreferrer" target="_blank">10.128.0.16:1338</a> <<a href="http://10.128.0.16:1338" rel="noreferrer" target="_blank">http://10.128.0.16:1338</a>>: getsockopt:<br>
>     connection refused<br>
>     4 times in the last minute<br>
>     6:39:45 PMWarningUnhealthy Liveness probe failed: Get<br>
>     <a href="https://10.128.0.16:1338/healthz" rel="noreferrer" target="_blank">https://10.128.0.16:1338/healthz</a> <<a href="https://10.128.0.16:1338/healthz" rel="noreferrer" target="_blank">https://10.128.0.16:1338/healthz</a>>:<br>
>     dial tcp <a href="http://10.128.0.16:1338" rel="noreferrer" target="_blank">10.128.0.16:1338</a> <<a href="http://10.128.0.16:1338" rel="noreferrer" target="_blank">http://10.128.0.16:1338</a>>: getsockopt:<br>
>     connection refused<br>
>     2 times in the last<br>
> <br>
>     and of course, when I browse the catalog, I don't see APB playbooks :-(<br>
> <br>
> <br>
>     On Mon, Aug 13, 2018 at 6:40 PM, David Zager <<a href="mailto:dzager@redhat.com" target="_blank">dzager@redhat.com</a><br>
>     <mailto:<a href="mailto:dzager@redhat.com" target="_blank">dzager@redhat.com</a>>> wrote:<br>
> <br>
>         An update to the liveness and readiness probes for 3.10 may be<br>
>         appropriate in openshift-ansible.<br>
> <br>
>         We are intentionally keeping the `dh-automation-broker-apb` out<br>
>         of the list of services. I suspect the issue here is simply with<br>
>         the log level of "not adding spec dh-automation-broker-apb to<br>
>         list of services due to error transforming to service - Could<br>
>         not find the parameter type for: "?<br>
> <br>
>         On Mon, Aug 13, 2018 at 12:35 PM Charles Moulliard<br>
>         <<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a> <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>>> wrote:<br>
> <br>
>             To get the pod up and running, I have removed the<br>
>             threshold's parameters from the APB's deployment resource<br>
>             ->liveness and readinessProbes.<br>
> <br>
>             But I see this error now<br>
> <br>
>             time="2018-08-13T16:30:09Z" level=info msg="Broker<br>
>             successfully bootstrapped on startup"<br>
>             time="2018-08-13T16:30:09Z" level=info msg="Listening on<br>
>             https://[::]:1338"<br>
>             time="2018-08-13T16:30:09Z" level=info msg="Ansible Service<br>
>             Broker Starting"<br>
>             ERROR: logging before flag.Parse: I0813 16:30:09.600491     <br>
>               1 serve.go:89] Serving securely on [::]:1338<br>
>             time="2018-08-13T16:30:20Z" level=info<br>
>             msg="AnsibleBroker::Catalog"<br>
>             time="2018-08-13T16:30:20Z" level=error msg="not adding spec<br>
>             dh-automation-broker-apb to list of services due to error<br>
>             transforming to service - Could not find the parameter type<br>
>             for: "<br>
>             10.128.0.1 - - [13/Aug/2018:16:30:20 +0000] "GET<br>
>             /ansible-service-broker/v2/catalog HTTP/1.1" 200 273918<br>
>             time="2018-08-13T16:30:40Z" level=info<br>
>             msg="AnsibleBroker::Catalog"<br>
>             time="2018-08-13T16:30:40Z" level=error msg="not adding spec<br>
>             dh-automation-broker-apb to list of services due to error<br>
>             transforming to service - Could not find the parameter type<br>
>             for: "<br>
>             10.128.0.1 - - [13/Aug/2018:16:30:40 +0000] "GET<br>
>             /ansible-service-broker/v2/catalog HTTP/1.1" 200 273918<br>
> <br>
>             On Fri, Aug 10, 2018 at 8:31 AM, Georgios Andrianakis<br>
>             <<a href="mailto:gandrian@redhat.com" target="_blank">gandrian@redhat.com</a> <mailto:<a href="mailto:gandrian@redhat.com" target="_blank">gandrian@redhat.com</a>>> wrote:<br>
> <br>
>                 Hello,<br>
> <br>
>                 I am seeing the same exact error as Charles under the<br>
>                 same conditions -> installing ASB on Openshift Origin 3.10.<br>
> <br>
>                 Regards<br>
> <br>
>                 On Thu, Aug 9, 2018 at 8:36 PM David Zager<br>
>                 <<a href="mailto:dzager@redhat.com" target="_blank">dzager@redhat.com</a> <mailto:<a href="mailto:dzager@redhat.com" target="_blank">dzager@redhat.com</a>>> wrote:<br>
> <br>
>                     After reading through the logs and seeing the way<br>
>                     the cut off, I'm curious if it's related to the<br>
>                     `initialDelaySeconds` in the broker's deployment<br>
>                     config<br>
>                     <<a href="https://github.com/openshift/openshift-ansible/blob/master/roles/ansible_service_broker/templates/asb_dc.yaml.j2#L59-L72" rel="noreferrer" target="_blank">https://github.com/openshift/openshift-ansible/blob/master/roles/ansible_service_broker/templates/asb_dc.yaml.j2#L59-L72</a>>.<br>
>                     With the upstream broker-apb we increased this from<br>
>                     15 seconds to 120<br>
>                     <<a href="https://github.com/openshift/ansible-service-broker/blob/master/apb/defaults/main.yml#L17" rel="noreferrer" target="_blank">https://github.com/openshift/ansible-service-broker/blob/master/apb/defaults/main.yml#L17</a>>.<br>
>                     I suspect that could be the cause.<br>
> <br>
> <br>
>                     On Thu, Aug 9, 2018 at 11:37 AM Jason Montleon<br>
>                     <<a href="mailto:jmontleo@redhat.com" target="_blank">jmontleo@redhat.com</a> <mailto:<a href="mailto:jmontleo@redhat.com" target="_blank">jmontleo@redhat.com</a>>><br>
>                     wrote:<br>
> <br>
>                         I made some comments in irc this morning.<br>
> <br>
>                         Our downstream CI test are passing as of this<br>
>                         morning using<br>
>                         openshift-ansible with downstream images. From<br>
>                         his gist logs it looked<br>
>                         like the broker started, but I am not clear on<br>
>                         what's happening to cause<br>
>                         the getsockopt error.<br>
> <br>
>                         <a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a><br>
>                         <<a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a>><br>
> <br>
>                         It seems like the service name resolved, but it<br>
>                         couldn't connect at the<br>
>                         address "172.30.206.144". Perhaps there is a<br>
>                         misconfiguration with the<br>
>                         docker or openshift networking?<br>
> <br>
>                         On 08/09/2018 11:30 AM, John Matthews wrote:<br>
>                          > +Jason/David who may be able to help<br>
>                          ><br>
>                          ><br>
>                          > Hi Charles,<br>
>                          ><br>
>                          > "I don't see anymore the apb-etcd pod wuth<br>
>                         openshift 3.10, it is normal ?"<br>
>                          ><br>
>                          > Yes, this is normal for 3.10, we moved to<br>
>                         CRDs for the data we store, so<br>
>                          > we dropped our own etcd and rely on regular<br>
>                         k8s API.<br>
>                          ><br>
>                          ><br>
>                          > "Error fetching catalog. Error getting broker<br>
>                         catalog:<br>
>                          ><br>
>                         Gethttps://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog<br>
> <br>
>                          ><br>
>                         <<a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a><br>
>                         <<a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a>>>:<br>
> <br>
>                          > dial tcp172.30.206.144:1338<br>
>                         <<a href="http://172.30.206.144:1338/" rel="noreferrer" target="_blank">http://172.30.206.144:1338/</a>>: getsockopt:<br>
>                          > no route to host"<br>
>                          ><br>
>                          > Please double check the namespace the<br>
>                         automation broker is running in.<br>
>                          > I wonder if there is a bug we missed with<br>
>                         "openshift-ansible" deploying<br>
>                          > with origin.<br>
>                          > I believe the majority of our QE scenarios<br>
>                         are with openshift-ansible<br>
>                          > with downstream OCP or 'oc cluster up' with<br>
>                         origin.<br>
>                          ><br>
>                          > Guessing, but perhaps there is a mismatch<br>
>                         between actual namespace<br>
>                          > broker is in versus what we are registering<br>
>                         to service-catalog.<br>
>                          ><br>
>                          ><br>
>                          ><br>
>                          ><br>
>                          > On Thu, Aug 9, 2018 at 4:29 AM, Charles<br>
>                         Moulliard <<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a><br>
>                         <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>><br>
>                          > <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a><br>
>                         <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>>>> wrote:<br>
>                          ><br>
>                          >     Additional info. The ClusterServiceBroker<br>
>                         reports this error<br>
>                          ><br>
>                          >     Error fetching catalog. Error getting<br>
>                         broker catalog: Get<br>
>                          ><br>
>                         <a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a><br>
>                         <<a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a>><br>
>                          >   <br>
>                           <<a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a> <<a href="https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog" rel="noreferrer" target="_blank">https://asb.openshift-ansible-service-broker.svc:1338/ansible-service-broker/v2/catalog</a>>>:<br>
>                          >     dial tcp <a href="http://172.30.206.144:1338" rel="noreferrer" target="_blank">172.30.206.144:1338</a><br>
>                         <<a href="http://172.30.206.144:1338" rel="noreferrer" target="_blank">http://172.30.206.144:1338</a>><br>
>                         <<a href="http://172.30.206.144:1338" rel="noreferrer" target="_blank">http://172.30.206.144:1338</a>>:<br>
>                          >     getsockopt: no route to host<br>
>                          ><br>
>                          >     On Thu, Aug 9, 2018 at 10:22 AM, Charles<br>
>                         Moulliard<br>
>                          >     <<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a><br>
>                         <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>><br>
>                         <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a><br>
>                         <mailto:<a href="mailto:cmoullia@redhat.com" target="_blank">cmoullia@redhat.com</a>>>> wrote:<br>
>                          ><br>
>                          >         Hi<br>
>                          ><br>
>                          >         I can't install ASB on origin 3.10<br>
>                         anymore using the playbook<br>
>                          >       <br>
>                           "openshift-ansible/playbooks/openshift-service-catalog/config.yml"<br>
>                          ><br>
>                          >         The apb deployment starts, a pod is<br>
>                         created and after x minutes<br>
>                          >         I see the status = "Failed" but the<br>
>                         pod's log doesn't report ERROR<br>
>                          ><br>
>                          >         Log :<br>
>                          ><br>
>                         <a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a><br>
>                         <<a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a>><br>
>                          >       <br>
>                           <<a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a> <<a href="https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb" rel="noreferrer" target="_blank">https://gist.github.com/cmoulliard/716468a9e9beb21ba860806a30e3ddcb</a>>><br>
>                          ><br>
>                          >         How can we troubleshoot and resolve<br>
>                         such error ?<br>
>                          >         I don't see anymore the apb-etcd pod<br>
>                         wuth openshift 3.10, it is<br>
>                          >         normal ?<br>
>                          ><br>
>                          >         Regards<br>
>                          ><br>
>                          >         Charles<br>
>                          ><br>
>                          ><br>
>                          ><br>
>                          >   <br>
>                           _______________________________________________<br>
>                          >     Ansible-service-broker mailing list<br>
>                          > <a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a><br>
>                         <mailto:<a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a>><br>
>                          >     <mailto:<a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a><br>
>                         <mailto:<a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a>>><br>
>                          ><br>
>                         <a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a><br>
>                         <<a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a>><br>
>                          >   <br>
>                           <<a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a> <<a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a>>><br>
>                          ><br>
>                          ><br>
> <br>
>                         -- <br>
>                         Jason Montleon     | email: <a href="mailto:jmontleo@redhat.com" target="_blank">jmontleo@redhat.com</a><br>
>                         <mailto:<a href="mailto:jmontleo@redhat.com" target="_blank">jmontleo@redhat.com</a>><br>
>                         Software Engineer  | gpg key: 0x069E3022<br>
>                         Red Hat, Inc.      | irc: jmontleo<br>
>                         desk: 978-392-3930 <tel:%28978%29%20392-3930> |<br>
>                         cell: 508-496-0663 <tel:%28508%29%20496-0663><br>
> <br>
>                     _______________________________________________<br>
>                     Ansible-service-broker mailing list<br>
>                     <a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a><br>
>                     <mailto:<a href="mailto:Ansible-service-broker@redhat.com" target="_blank">Ansible-service-broker@redhat.com</a>><br>
>                     <a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a><br>
>                     <<a href="https://www.redhat.com/mailman/listinfo/ansible-service-broker" rel="noreferrer" target="_blank">https://www.redhat.com/mailman/listinfo/ansible-service-broker</a>><br>
> <br>
> <br>
> <br>
> <br>
<br>
-- <br>
Jason Montleon     | email: <a href="mailto:jmontleo@redhat.com" target="_blank">jmontleo@redhat.com</a><br>
Software Engineer  | gpg key: 0x069E3022<br>
Red Hat, Inc.      | irc: jmontleo<br>
desk: 978-392-3930 | cell: 508-496-0663<br>
</blockquote></div></div>