Hi James and Dan,

I restarted sdnc pod to enable karaf log, after the restart sdnc healthcheck 
passed. I don’t know if you or someone else happened to make some change, or 
restart helped.  Similarly after I restarted APPC pod, its healthcheck passed 
as well. This is Casablanca released version, we will make a note of it to see 
if it happens again.

Thanks for your help,

-Yang


From: [email protected] [mailto:[email protected]] On 
Behalf Of Yang Xu
Sent: Wednesday, December 12, 2018 2:43 PM
To: [email protected]; [email protected]; TIMONEY, DAN 
<[email protected]>
Subject: Re: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Hi James,

Thanks for the reply. I don’t know much about SDNC cluster settings. On SB05, 
from Robot container I can ping sdnc-cluster.onap and also I can telenet 
sdnc.onap 8282 port. I am guessing, port 8181 is only used internally by SDNC 
domain to check heartbeat of each replica, so port 8181 doesn’t need to be 
exposed as a service port. Please let me if you need more info about SB05 
testbed.

Regards,
-Yang

From: [email protected]<mailto:[email protected]> 
[mailto:[email protected]] On Behalf Of James MacNider
Sent: Wednesday, December 12, 2018 1:30 PM
To: Yang Xu (Yang, Fixed Network) 
<[email protected]<mailto:[email protected]>>; TIMONEY, DAN 
<[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>
Subject: Re: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

It’s possible there’s a misconfiguration here, but I’ve not had the chance to 
investigate yet.

One thing to keep in mind is that when we’re discussing “domains” in the 
context of a Kubernetes cluster, what you’re actually referring to is the name 
of a Kubernetes service that is used to load-balance across one or more pods.  
These names are tied to specific ports that the pods expose.  In the case of 
SDN-C, these are the service names that are being exposed, along with their 
associated ports:

$ kubectl get services -n onap
NAME                               TYPE        CLUSTER-IP      EXTERNAL-IP   
PORT(S)                                                       AGE
…
controller-blueprints              ClusterIP   10.43.14.62     <none>        
8080/TCP                                                      16m
controller-blueprints-db           ClusterIP   None            <none>        
3306/TCP                                                      16m
…
sdnc                               NodePort    10.43.161.171   <none>        
8282:30202/TCP,8202:30208/TCP,8280:30246/TCP,8443:30267/TCP   16m
sdnc-ansible-server                ClusterIP   10.43.83.83     <none>        
8000/TCP                                                      16m
sdnc-cluster                       ClusterIP   None            <none>        
2550/TCP                                                      16m
sdnc-dbhost                        ClusterIP   None            <none>        
3306/TCP                                                      16m
sdnc-dbhost-read                   ClusterIP   10.43.51.29     <none>        
3306/TCP                                                      16m
sdnc-dgbuilder                     NodePort    10.43.50.6      <none>        
3000:30203/TCP                                                16m
sdnc-dmaap-listener                ClusterIP   None            <none>        
<none>                                                        16m
sdnc-portal                        NodePort    10.43.185.129   <none>        
8843:30201/TCP                                                16m
sdnc-sdnctldb01                    ClusterIP   None            <none>        
3306/TCP                                                      16m
sdnc-sdnctldb02                    ClusterIP   None            <none>        
3306/TCP                                                      16m
sdnc-ueb-listener                  ClusterIP   None            <none>        
<none>                                                        16m

So if the cluster health check is using port 8181, this could a problem, as I 
don’t’ see any services exposing that port…




From: Yang Xu (Yang, Fixed Network) 
<[email protected]<mailto:[email protected]>>
Sent: Wednesday, December 12, 2018 12:43 PM
To: TIMONEY, DAN <[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>; James MacNider 
<[email protected]<mailto:[email protected]>>
Subject: RE: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Dan,

I see your point. From sdnc oom script, I see sdnc-cluster.onap is used for 
SDNC cluster health check and it uses port 8181. The issue I reported is from 
SDNC component health check, the port used is 8282. Maybe both 
sdnc-cluster.onap and sdnc.onap are valid.

Thanks,
-Yang

From: TIMONEY, DAN [mailto:[email protected]]
Sent: Wednesday, December 12, 2018 9:59 AM
To: Yang Xu (Yang, Fixed Network) 
<[email protected]<mailto:[email protected]>>; 
[email protected]<mailto:[email protected]>; MACNIDER, 
JAMES <[email protected]<mailto:[email protected]>>
Subject: Re: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Yang,

The problem is that the scripts that setup clustering rely on the domain 
sdnc-cluster.onap being setup.  I just rebuilt my WindRiver test ONAP 
environment last week, following the instructions in the OOM User Guide for 
installing with Rancher, and using the casablanca branch of OOM.  In my 
environment, the pod names and domain name work as they should.

If we're saying the domain name should be sdnc.onap instead of 
sdnc-cluster.onap, I'd be okay with making that change (which is really in 
OOM).  However, the sdnc.onap domain also isn't working between pods  (i.e. if 
I exec to dev-sdnc-sdnc-0,  I can't get to dev-sdnc-sdnc-1.sdnc.onap either).  
Also, I notice in my local environment that I can ping 
dev-sdnc-1.sdnc-cluster.onap from dev-sdnc-0, but not dev-sdnc-1.sdnc.onap ... 
so I think the domain 'sdnc-cluster.onap' is correct, but for some reason not 
working.


James – can you perhaps take a look at SB05?  I’m not really familiar enough 
with the details of how that sdnc-cluster.onap domain is set up within K8S to 
be of too much more help here other than to point out it’s not working.

Dan

Dan Timoney
Principal Technical Staff Member
AT&T
Email : [email protected]<mailto:[email protected]>
Office : +1 (732) 420-3226
Mobile : +1 (201) 960-1211
200 S Laurel Ave, Rm E2-2A03
Middletown, NJ 08873

From: "Yang Xu (Yang, Fixed Network)" 
<[email protected]<mailto:[email protected]>>
Date: Tuesday, December 11, 2018 at 11:43 PM
To: "TIMONEY, DAN" <[email protected]<mailto:[email protected]>>, onap-discuss 
<[email protected]<mailto:[email protected]>>
Subject: RE: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Hi Dan,

I really appreciate you looked into the issue. The healthcheck request is sent 
from ONAP Robot docker, and from Robot report log I can see the domain is 
sdnc.onap which can be resolved when testing with telnet from Robot. In fact, 
when I try to access SDNC API page with 
http://sdnc_external_ip:30202/apidoc/explorer/index.html<https://urldefense.proofpoint.com/v2/url?u=http-3A__sdnc-5Fexternal-5Fip-3A30202_apidoc_explorer_index.html&d=DwMGaQ&c=LFYZ-o9_HUMeMTSQicvjIg&r=Ms-DNhR23dO4sQFfYRm4ow&m=Br1l2AklrgZQToyKfpJ6ktBL8xbsbvxuEt3_u21uEfc&s=RquV8JPmf7KUk71vDRnba_rhntyrvyjRNckhBaC3TSc&e=>,
 it also fails with 404. The same url works for an ONAP deployment from last 
week.

Regards,
-Yang



From: TIMONEY, DAN [mailto:[email protected]]
Sent: Tuesday, December 11, 2018 6:27 PM
To: [email protected]<mailto:[email protected]>; Yang Xu 
(Yang, Fixed Network) <[email protected]<mailto:[email protected]>>
Subject: Re: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Yang,

Sorry, I meant to explain a little more  …..

We use the sdnc-cluster.onap domain names in our akka.conf file to set up 
clustering.   So, if that domain isn’t working – as seems to be the case in 
SB05 – the nodes won’t be able to cluster properly.  That’s what’s causing that 
404 error.

So the root cause of this problem is that the sdnc-cluster.onap domain isn’t 
resolving properly.  As I mentioned, that’s actually working for me in my 
windriver env, so I suspect the error is probably some environment error.

Dan

--
Dan Timoney
SDN-CP Development
ONAP Project Technical Lead : CCSDK and SDNC

Please go to  D2 ECOMP Release Planning 
Wiki<https://wiki.web.att.com/display/DERP/D2+ECOMP+Release+Planning+Home> for 
D2 ECOMP Project In-take, 2016 Release Planning, Change Management, and find 
key Release Planning Contact Information.


From: "TIMONEY, DAN" <[email protected]<mailto:[email protected]>>
Date: Tuesday, December 11, 2018 at 6:21 PM
To: onap-discuss 
<[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>
Subject: Re: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Yang,

I replied to this on RocketChat as well, but there’s some environmental issue 
going on in that environment.

In my local k8s env, based on the latest OOM casablanca branch in gerrit, my 
pods are named dev-sdnc-0, dev-sdnc-1 and dev-sdnc-2, and K8S defines an 
‘sdnc-cluster.onap’ domain where the nodes can refer to each other.  In other 
words, if I exec into dev-sdnc-0,  I can then telnet to 
dev-sdnc-1.sdnc-cluster.onap on port 2550 to confirm akka connectivity.

In SB05, the sdnc node names have the ‘sdnc’ part doubled (e.g. 
dev-sdnc-sdnc-0), and the sdnc-cluster.onap domain is not working (i.e. telnet 
dev-sdnc-sdnc-1.sdnc-cluster.onap tells me name/service not known).

I think this is some sort of environment issue, maybe with the environment.yaml 
file for that environment.  In my local k8s env, I’m just using the out of the 
box OOM helm charts (only change I make is in onap/values.yaml to disable the 
components I don’t need).

Dan

--
Dan Timoney
SDN-CP Development
ONAP Project Technical Lead : CCSDK and SDNC

Please go to  D2 ECOMP Release Planning 
Wiki<https://wiki.web.att.com/display/DERP/D2+ECOMP+Release+Planning+Home> for 
D2 ECOMP Project In-take, 2016 Release Planning, Change Management, and find 
key Release Planning Contact Information.


From: onap-discuss 
<[email protected]<mailto:[email protected]>> on behalf of 
Yang Xu <[email protected]<mailto:[email protected]>>
Reply-To: onap-discuss 
<[email protected]<mailto:[email protected]>>, 
"[email protected]<mailto:[email protected]>" 
<[email protected]<mailto:[email protected]>>
Date: Tuesday, December 11, 2018 at 1:31 PM
To: onap-discuss 
<[email protected]<mailto:[email protected]>>
Subject: [onap-discuss] [sdnc][appc][integration] SDNC and APPC fail on 
healthcheck with 404

Hi Dan and Taka,

I deployed ONAP Casablanca on SB-05 in Windriver lab and found both SDNC and 
APPC failed their healthcheck with 404 error. SDNC and APPC pods seem ok and 
other ONAP modules passed their healthcheck. Please help take a look.

Regards,
-Yang

This email and the information contained herein is proprietary and confidential 
and subject to the Amdocs Email Terms of Service, which you may review at 
https://www.amdocs.com/about/email-terms-of-service


-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#14459): https://lists.onap.org/g/onap-discuss/message/14459
Mute This Topic: https://lists.onap.org/mt/28722186/21656
Group Owner: [email protected]
Unsubscribe: https://lists.onap.org/g/onap-discuss/unsub  
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to