Created attachment 1360185 [details] console logs Description of problem: ======================= My setup comprised of beaker (perf) machines, for storage as well as tendrl server. 'ip a' on the tendrl server showed the below output. Even after specifying hostname in site.yml file, 'etcd_ip_address' was getting udpated to a different IP -- and that resulted in failure while creating tendrl-admin-user. Pasted below is the error output seen on console. I suppose we need better resolution of IP address in the line: etcd_ip_address: "{{ lookup('dig', ansible_fqdn) }}" TASK [tendrl-ansible.tendrl-server : Create Tendrl admin user] ******************** ^[[0;31mfatal: [gqas012.sbu.lab.eng.bos.redhat.com]: FAILED! => {"changed": true, "cmd": "RACK_ENV=production rake etcd:load_admin", "delta": "0:00:00.606007", "end": "2017-11-28 20:49:48.948206", "failed": true, "failed_when_result": true, "rc": 1, "start": "2017-11-28 20:49:48.342199", "stderr": "rake aborted!\nConnection refused - connect(2)\n\nTasks: TOP => etcd:load_admin\n(See full trace by running task with --trace)", "stderr_lines": ["rake aborted!", "Connection refused - connect(2)", "", "Tasks: TOP => etcd:load_admin", "(See full trace by running task with --trace)"], "stdout": "\"Generating default Tendrl admin\"", "stdout_lines": ["\"Generating default Tendrl admin\""]}^[[0m 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN qlen 1 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo valid_lft forever preferred_lft forever inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: em1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000 link/ether 78:2b:cb:74:46:ac brd ff:ff:ff:ff:ff:ff inet 10.16.156.33/21 brd 10.16.159.255 scope global dynamic em1 valid_lft 57514sec preferred_lft 57514sec inet6 fe80::7a2b:cbff:fe74:46ac/64 scope link valid_lft forever preferred_lft forever 3: em2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN qlen 1000 link/ether 78:2b:cb:74:46:ad brd ff:ff:ff:ff:ff:ff 4: p2p1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 qdisc mq state UP qlen 1000 link/ether 90:e2:ba:07:a2:84 brd ff:ff:ff:ff:ff:ff inet 192.168.96.143/24 brd 192.168.96.255 scope global p2p1 valid_lft forever preferred_lft forever inet6 fe80::92e2:baff:fe07:a284/64 scope link valid_lft forever preferred_lft forever 5: p2p2: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN qlen 1000 link/ether 90:e2:ba:07:a2:85 brd ff:ff:ff:ff:ff:ff Version-Release number of selected component (if applicable): ============================================================ tendrl-ansible-1.5.4-2.el7rhgs.noarch How reproducible: ================= 1:1 Steps to Reproduce: =================== 1. Have storage nodes and a rhel7 machine for web-admin server, which has 2 IP addresses. 2. 'Run ansible-playbook -i hosts site.yml' from the tendrl-server Actual results: =============== Step 2 fails in the middle Expected results: ================ tendrl-ansible should succeed in resolving and picking the right IP address. Additional info: ================ PFA the console logs.
I need to review this. Reply is expected no sooner than on Monday.
Since workaround available, moving this out of this release.
Sorry, I didn't have the time to try to replicate this. That said, is this a issue caused by (previously) unclear downstream documentation or actual issue which one would hit even when follows upstream installation docs and README file from rpm package? Context: The docs states one needs to update variables in example playbook file when needed. The defaults there are not expected to just work in all circumstances.
Fixing typo: (In reply to Martin Bukatovic from comment #6) > upstream installation docs and README file from rpm package? s/upstream/downstream/
Remove 'and not in the tendrl-ansible configuration file'
Updated,please check now
IP address of a machine can change any time unless we have a static ip. If there is a change in ip, we may need to update wherever the system has configured to use the ip. So its better to use static-ip or host-fqdn which can be configured to resolve the ip at top level. @Sweta Anandpara Could you please post the results of the following output: $ip a $nslookup <hostname>
Ok, That means even ping <hostname/fqdn> might return incorrect ip. I hope this may not be an ansible or tendrl-ansible issue. It could be a dns issue and for some reason (dnscache) the host could unable to provide the correct ip address of the given fqdn. I will try to reproduce and update here and come up with a possible solution. @Sweta, did you installed etcd in a different server or in a same server where you have installed tendrl-api and tendrl-ui?
I'm verifying this Bug based on following points: * All mandatory variables have to be configured manually by user, no automated guess is performed, so it is up to the user, to properly configure the variables. * If some mandatory variable is not configured, execution of tendrl-ansible site.yml playbook is failed with proper error message: You need to define all mandatory ansible variables to run this playbook, see README file for guidance. * The required mandatory variables are described in the README and will be described in documentation (Bug 1584737). Also this bug doesn't cover case, when Gluster Cluster is using different network than the monitoring (similar scenario is covered by Bug 1573075). Tested with: tendrl-ansible-1.6.3-5.el7rhgs.noarch >> VERIFIED
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHSA-2018:2616
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days