Hide Forgot
Created attachment 1160872 [details] sosreport Description of problem: Node ng can not register to satellite through cockpit. Version-Release number of selected component (if applicable): rhev-hypervisor7-ng-3.6-20160518.0 imgbased-0.6-0.1.el7ev.noarch cockpit-0.103-1.el7.x86_64 cockpit-ovirt-dashboard-0.9.1-0.0.1.el7ev.noarch How reproducible: 100% Steps to Reproduce: 1. Install node ng 2. Install the rpm from satellite to provide CA files. 3. In subscriptions page, file the satellite url in the customer url field. And alos file the username,passowrd. 4. Click register Actual results: 1. It reports "Unable to reach the server at *****:443/subscription". Expected results: 1. It should register to satellite succeed. Additional info:
(In reply to Stef Walter from comment #1) > Do you have a test system we can reproduce this bug against? Hi Stef, Which one your need for testing, the node ng or satellite6 server?
(In reply to Stef Walter from comment #3) > The satellite6 server would be needed. Thanks. Hi Stef, Due to our policy, I can only share you the env through the email. So I send you a email to share you the satellite6 server's info. Thanks! huiwa
Other bugs cover activation key, proxy and org support. Also, we have upsteam CI coverage via candlepin starting 0.114. It is now difficult to tell where this problem came from, or if it even came from Cockpit, and if there are no objections I'll close this bug.
Chen and Wei, could you help on comment 5?
Hi Stef, After confirming with satellite team, you can try with port 80 for listening instead of port 443.
Test Version redhat-virtualization-host-4.0-20160810.1.x86_64.liveimg.squashfs imgbased-0.8.3-0.1.el7ev.noarch Test steps: 1. Install RHVH with kickstart file in attachment. 2. Install the CA files 3. In node console shell, input the below command: #subscription-manager register --serverurl=<satellite server> --username=username --password=passwd 4.Check the result Result: "Unable to reach the server at satellite61.redhat.com:443/subscription" is displayed on the console. QE can reproduce this bug with the latest RHVH version.
There is still a chance that this is a bug on the Node side. Douglas, can you reproduce this bgu with our latest RHVH 4 build?
(In reply to weiwang from comment #10) > Test Version > redhat-virtualization-host-4.0-20160810.1.x86_64.liveimg.squashfs > imgbased-0.8.3-0.1.el7ev.noarch > > Test steps: > 1. Install RHVH with kickstart file in attachment. > 2. Install the CA files > 3. In node console shell, input the below command: > #subscription-manager register --serverurl=<satellite server> > --username=username --password=passwd > 4.Check the result > > Result: > "Unable to reach the server at satellite61.redhat.com:443/subscription" is > displayed on the console. > > QE can reproduce this bug with the latest RHVH version. Could you please share the satellite test env and CA files to install with me by email as well? Thanks!
(In reply to Douglas Schilling Landgraf from comment #12) > (In reply to weiwang from comment #10) > > Test Version > > redhat-virtualization-host-4.0-20160810.1.x86_64.liveimg.squashfs > > imgbased-0.8.3-0.1.el7ev.noarch > > > > Test steps: > > 1. Install RHVH with kickstart file in attachment. > > 2. Install the CA files > > 3. In node console shell, input the below command: > > #subscription-manager register --serverurl=<satellite server> > > --username=username --password=passwd > > 4.Check the result > > > > Result: > > "Unable to reach the server at satellite61.redhat.com:443/subscription" is > > displayed on the console. > > > > QE can reproduce this bug with the latest RHVH version. > > Could you please share the satellite test env and CA files to install with > me by email as well? > > Thanks! Ok, no problem! The email has been sent to you, please check!
Is there any progress? I'm tempted to close this issue, as the case is covered by upstream Cockpit CI. Being unable to reach a server hints at an underlying issue that isn't related to Cockpit.
In the logs I see: 2016-05-23 03:26:11,256 [ERROR] rhsmcertd-worker:16736 @rhsmcertd-worker:47 - Either the consumer is not registered or the certificates are corrupted. Certificate update using daemon failed. Could you please try to reproduce this issue on plain RHEL with the exact same steps you used on RHVH?
Test Version: redhat-release-server-7.2-7.el7.x86_64 cockpit-0.114-2.el7.x86_64 cockpit-ws-0.114-2.el7.x86_64 Test Steps: 1. Install vm with RHEL-7.2-20151015.0-Server-x86_64-dvd1.iso 2. Install cockpit rpm 3. Enable cockpit services 4. Install the CA files 5.a). In node console shell, input the below command: #subscription-manager register --serverurl=<satellite server> --username=username --password=passwd 6.a). Check the result 5.b-1). Login cockpit website hostIP:9090 with root account b-2). In subscriptions page, file the satellite url in the customer url field. And also file the username,passowrd, then click "register" 6.b). Check the display on node screen. Result: "Unable to reach the server at satellite61.redhat.com:443/subscription" is displayed on screen. More info: I also test with satellite 6.2 server, the result is the same.
Dominik is away. The problem is that I see this: $ nslookup satellite61.redhat.com Server: 10.38.5.26 Address: 10.38.5.26#53 ** server can't find satellite61.redhat.com: NXDOMAIN Is there a server against which we can reproduce this bug?
I have tried to reproduce the report and got the same error from Stef in comment#17. weiwang, do you have any advise? Thanks!
(In reply to Douglas Schilling Landgraf from comment #18) > I have tried to reproduce the report and got the same error from Stef in > comment#17. > > weiwang, do you have any advise? > > Thanks! We are doubt about the satellite server have some problem, RHEL7 and RHVH4.0 cannot register satellite 6.2 successful, so we now debug this. I will update the info later.
After retry and research this bug again, we can register to satellite from cockpit webUI and CLI successful. The correct point is the server Url need to be <satellite hostname>/rhsm. I have retest with satellite6.2+RHEL7.2 using cockpit webUI and CLI, the bug cannot be reproduced.
Dominik, should we add /rhsm automatically to a satellite URL without a path?
Technically we don't know it's a satellite system, upstream we support candlepin. The suffix is entirely up to the server's configuration. We can discuss trying the suffix automatically if we can't connect using just the hostname.