SMA-SM - Cannot login to the installation portal

After deploying the CDF core content I cannot login to the Installation Portal.

When I load the URL: the login page appears.

I can fill the userid and password fields, but when I click on login the browser stucks in Loading ..... and nothing heapens.

The pods are started succesfully and I can't see anything wrong in the logs.

Could someone tell me what could be the problem?

  • I have noticed that reboot sometime helps in this case. Has twice in SMAX install same issue.

  • Hello Misaq,

    Thanks for you answer!

    I've rebooted the system but the situation is the same.

    The question is actually that how can I check if the install frontend is functioning correctly?

    I checked the POD's log files, but nothing special in them.

  • Unfortunately I have not experineced this yet. But for troubleshooting may be following would help:

    • check   output
    • check kube-start.log (might be a dufferent name(its in either /tmp or under kubernetes folder).
    • try different browser.
    • "loading..." issues I had were mostly in 2018.02 and 2018.05 CDF  after launching the installation portal but only once I had to re-install otherwise reboot helped.
    • If you look in the log folder under /opt/kubernetes    one of them should be indicating the error
    • Also follow the troubleshooting section :
  • I've rebuilt it from scrach and still the same.

    But the source of the problem should be this:

    E0115 14:31:45.048841 4879 proxier.go:584] Failed to execute iptables-restore for nat: exit status 1 (iptables-restore: invalid option -- '5'
    Try `iptables-restore -h' for more information.
    E0115 14:31:45.055563 4879 proxier.go:610] Failed to execute iptables-restore for filter: exit status 1 (iptables-restore: invalid option -- '5'
    Try `iptables-restore -h' for more information.
    E0115 14:31:45.069480 4879 proxier.go:792] Failed to execute iptables-restore for nat: exit status 1 (iptables-restore: invalid option -- '5'

    One thing that is more interesting: I have a conrol machine with CentOS (where everithing works) and with the same version of iptables-restore v1.4.21

  • Have you triend disabling the firewall?(I am guessing you are in testing mode for now)

    I think its in the instruction.

    systemctl stop  firewalld

    systemctl disable firewalld

    also  disable selinux


    This should rule out the iptable issue. ALso if you are using browser from other client , can try locally on the server to test if something blocking it there.

  • Yes, I went through these things.

    Firewall disabled, selinux permissive and this is a minimal install without X, so I can not run local browsers.

    The frontend keeps tracking where we are and when I interrupt the configuration and load again the frontend, it knows where were we previously. Do you know where this information is stored?

  • Verified Answer

    I've got a solution from Support.

    We had to set the following permission in order to installer to be able to read the suite metadata:

    chmod 755 -R /opt/kubernetes/cfg/suite-metadata/

  • strange. I have this setting usually when Install. There might be something wrong with the installation process.