Installing Ops Center Administrator with the application installer

Ops Center Administrator Getting Started Guide
Version
10.9.x
File Size
782 KB
Audience
anonymous
Part Number
MK-99ADM000-17

You can install Ops Center Administrator in a Linux environment that is running a supported version of a container runtime.

To enable maximum control of the environment, the application installer does not include a container runtime, an operating system, or a VM.

  • Do not install multiple container runtimes such as Docker and Podman on the host OS. Always use a single container runtime.
  • If you want to register Ops Center Administrator with Ops Center Common Services, do the following:
    • Install Python3 before running the installer.
    • Make sure that the host name of Ops Center Common Services is resolvable from the Ops Center Administrator server. If you want to use a host name that is not an FQDN, set the IP address and the host name in the /etc/hosts file for name resolution.
    • After modifying the /etc/hosts file, run the following command to restart the Ops Center Administrator-related service.

      For the Docker environment, run:

      # systemctl restart docker

      For the Podman environment, run:

      # systemctl restart rainier
    • Make sure that you have a user account with Ops Center Common Services that has the "Application Administrator" role to run the script.

Verify the following:

  • You have root access or normal access to the OS where you plan to install Ops Center Administrator. If you log in as a normal user, use the sudo command to complete the procedure as the root user.
  • IP forwarding and br_netfilter for the IP V4 network is installed on the operating system.

    Verify by using the sysctl command (1 means enabled):

    # sysctl net.ipv4.ip_forward
net.ipv4.ip_forward = 1
# sysctl net.bridge.bridge-nf-call-iptables
net.bridge.bridge-nf-call-iptables = 1
  • The OS running firewalld (for example, RHEL 7 or later) is configured to allow communication between containers as follows:
    • Enable communication by adding the following to the trusted zone:
      • For Docker, add docker0.
      • For Podman add cni-podman0.
    • The following example is for Docker:
      # firewall-cmd --zone=trusted --change-interface=docker0 --permanent
      # firewall-cmd --reload
    • Enable IP masquerading for the default zone:
      # firewall-cmd --add-masquerade --permanent
      # firewall-cmd --reload
  • A supported version of container runtime is installed in a Linux environment.
    When using Docker:
    • Do not set the following service options to false:
      • icc: Enable inter-container communication (default true).
      • ip-forward: Enable net.ipv4.ip_forward (default true).
      • icc: Enable IP masquerading (default true).
      • iptables: Enable addition of iptables rules (default true).
    • If the FORWARD chain does not contain the DOCKER chain in the iptables, restart the Docker service by running the following command:
      # systemctl restart docker
    • The recommended settings for using the application installer are as follows:
      • If you use the JSON-file logging driver, set the maximum log size to 50 MiB and the maximum number of files to 3. 
         --log-opt max-size=50m --log-opt max-file=3
      • Set ExecReload. 
        ExecReload=/bin/kill -s HUP $MAINPID
      • Set Delegate to yes. 
        Delegate=yes
      • Set KillMode to process. 
        KillMode=process

      • Set Restart. 

        Restart=on-failure

        Refer to the example provided for general information about the Docker.service file.

    When using Podman:
    • If the supported version of Podman is not installed in the environment, you must configure Yellowdog Updater, Modified (YUM) settings to install packages over a network. The application installer connects to the configured YUM repository and installs the required version of Podman. The packages related to Podman are located in the latest BaseOS and AppStream repositories.

    • If you want to install or upgrade Podman yourself, you can run the following command: 

      yum install podman required-version

      For example, to install Podman v3.3.x: 

      yum install podman-3.3.*

      The asterisk indicates to obtain and install the latest patch version available in the repository.

      Note: Downgrading the Podman version is not supported.
    • If you install or upgrade Ops Center Administrator while suppressing the upgrade of Podman, the installation fails with the error - [Error] Failed to install Podman x.x.x from package repository. Confirm the network or repository server setting, and retry. Unlock the suppression and install or upgrade Ops Center Administrator again. After completing the installation, suppress the upgrade of Podman again.
      Note: You can either upgrade Podman during the installation or upgrade process, or you can do it yourself after installation. We recommend that you upgrade Podman during the installation process because you cannot upgrade Podman from 3.x to 4.x after installation without using the backup and restore procedure. If you must upgrade Podman after installation, create a backup file of your existing Ops Center Administrator instance, remove and reinstall Ops Center Administrator with a Podman upgrade, and then restore the backup.
    • If you install Podman 3.3.x, or upgrade Podman from 3.3.x to any version, or run any Podman command on the server using Podman 3.3.x, a warning message Failed to decode the keys [<key1>, <key2>, ..., <keyN>] from "/usr/share/containers/containers.conf" may appear. Ignore this message because it does not affect Ops Center Administrator.
    • If you cannot use YUM to install Podman because your management server is not connected to the network, you must get the Podman software from the OS media (ISO image or CD-ROM).

      For example, the minimum supported version of Podman 3.3.x is available with Red Hat Enterprise Linux and Oracle Linux version 8.5, Podman 4.0.x is available with version 8.6, and Podman 4.2.x is available with version 9.1. Therefore, regardless of the OS version that you are using, you must download the OS that includes the version of Podman that you want to use.

      1. Download the Linux ISO image (for example, redhat 8.5 iso).
      2. Mount the ISO image using the following command:

        mount /dev/cdrom /media

        For example: mount -o loop rhel-8.5-x86_64-dvd.iso /media

      3. If the /etc/yum.repos.d directory contains an existing repo file, rename the file extension or delete it.
      4. Create the yum repository file by running the following command:

        vim /etc/yum.repos.d/local.repo

      5. Add the required definition lines as shown in the following examples, and then save and close the file:

        For Oracle Linux

        [LocalRepo_BaseOS]name= LocalRepo_BaseOS
gpgcheck=0
enabled=1
baseurl=file:///media/BaseOS/
LocalRepo_AppStream]
name=LocalRepo_AppStream
gpgcheck=0
enabled=1
baseurl=file:///media/AppStream/

        For Red Hat Enterprise Linux

        [LocalRepo_BaseOS]
name=LocalRepo_BaseOS
metadata_expire=-1
enabled=1
gpgcheck=0
baseurl=file:///media/BaseOS/
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release
 
[LocalRepo_AppStream]
name=LocalRepo_AppStream
metadata_expire=-1
enabled=1
gpgcheck=0
baseurl=file:///media/AppStream/
gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-redhat-release
      6. Verify the required library by running the following command:
        yum repolist
      7. Install podman by using the following command :
        yum install podman-required-version

        For example, to install Podman v3.3.x: 

        yum install podman-3.3.*
  • The available space on the server is 100 GiB including temporary space. For details, refer to Minimum system requirements.
  • The server has 8 to 10 GiB of available RAM. For details, refer to Minimum system requirements.
    Note: Before starting the installation, as a best practice, do not install Ops Center Administrator in a location running other applications.
  1. In the Linux environment, configure the network interface that will access Ops Center Administrator.
    Ops Center Administrator supports user interface and API access by using an IPv4 address.
  2. Copy the tar file ops-center-administrator-xx.tar.gz from the installation media to any folder in the Linux environment and unzip it.
  3. Navigate to the unzipped folder and run install.sh.
    At the prompts, enter the following:
    1. Enter the username for the installer:

      Enter sysadmin

    2. Enter the user password:

      Enter sysadmin

    3. Enter host's IP:

      Enter the IP address for Ops Center Administrator. This IP address is also used for SNMP communications with the storage system.

    4. Enter the Service port number (HTTPS, default 443):

      Enter the service port for accessing Ops Center Administrator. The default service port is HTTPS, 443. You can proceed with the default service port or enter your own.

      Note: If you are using Ops Center Administrator with Ops Center Common Services or Ops Center Protector, you must enter a port other than the default (443), which causes a conflict. We suggest 20961.
    5. If you want to register Ops Center Administrator with Ops Center Common Services during installation, enter y at the prompt: 
      Do you wish to configure Ops Center [y/n]

      You are then prompted to enter a user name and password for Ops Center Common Services and the name and description of the Ops Center Administrator instance to register.

      Note: If both Ops Center Administrator and Ops Center Common Services are v10.9.0 or later, then the token authentication time-out configured in Ops Center Common Services is reflected in Ops Center Administrator automatically. This time-out configuration applies not only to users managed in Ops Center Common Services, but also to local users in Ops Center Administrator.

      For details, see the Portal Help for Ops Center Common Services.

      Note: In Ops Center Administrator, the Auto-refresh setting configured in the Ops Center Portal is ignored.
    Note: During installation, vm.max_map_count is set to 262144 in /etc/sysctl.conf.
    The installation may take a few minutes. At completion, messages indicate the following:
    • The application was successfully added.
    • The API is ready.
    • Any pre-existing app manager containers have been removed.
  4. Set the SNMP IP address in the virtual appliance manager tool:
    1. Open a browser and enter https://ip-address/vam in the address bar.
      The login credentials are sysadmin/sysadmin.
    2. Click Network Settings, then enter the SNMP IP address for the storage system.
  5. For Podman, suppress the upgrade of Podman to avoid unintentionally upgrading to the unsupported version.

    For example, you can use yum-plugin-versionlock or you can add the exclude parameter to the yum.conf file.

Troubleshooting the installation

If the installation fails, try the following:
  • If the installation fails with the (401) error code, the user name and password specified by the installer was incorrect. Retry the installation and ensure that the credentials used are sysadmin/sysadmin.
  • Check your container network configuration and make sure your container runtime is working properly. Container runtime requires its network interface (docker0 for Docker, or cni-podman0 for Podman) to be in a trusted zone in your operating system.
  • Check your YUM settings and the host network to make sure that your system can connect to the YUM repository.
  • If you use a local YUM mirror repository server, confirm the setting of the HTTP server and whether the repository data which is gathered by the reposync command exists correctly.
  • If the installation fails and an error message iptables: No chain/target/match by that name is output in the Ops Center Administrator installation log, remove Ops Center Administrator and restart the Docker service by running the command: 
    # systemctl restart docker

    Then, retry the installation.

  • Delete all Ops Center Administrator containers, images, and files and then start the installation again.
  • Check the Docker or Podman logs.

    Consult the documentation of your container runtime for more information on how to check logs.

  • Journal entries may have additional information about the error. To view the journal log, connect to the host with the root user or a normal user account and run the following commands:
    Note: If you log in as a normal user, use the sudo command to complete the following procedure as the root user.
    • Docker: journalctl --no-pager -u docker
    • Podman: journalctl --no-pager -u rainier

    If the issue persists, collect the installation log that was created under /var/logs/rainier-install and contact customer support.

  • If the installation produces any warnings, they may point to the cause of the problem. Correct any issues the installer identifies, delete any Ops Center Administrator containers and images, and start the installation again. You can log in as the root user or a normal user. If you log in as a normal user, use the sudo command to complete the following procedure as the root user.
    To remove files, run the following commands:

    rm -f /opt/rainier/bin/rainier-getlogs

    rm -f /opt/rainier/bin/rainier-replace-jdk

    To remove container images and containers that you do not manage, run these commands with the root account (use podman instead of docker depending on your container runtime):
    1. docker stop $(docker ps --format "{{.ID}} {{.Image}}" -a | grep "rdocker:6000/" | awk '{ print $1 }')
    2. docker rm -fv $(docker ps --format "{{.ID}} {{.Image}}" -a | grep "rdocker:6000/" | awk '{ print $1 }')
    3. docker rmi $(docker images --format "{{.ID}} {{.Repository}}" | grep "rdocker:6000/" | awk '{ print $1 }')
    4. docker volume rm nginx-certificates
    5. docker volume rm nginx-certificates-override
    6. docker volume rm nginx-confd
    7. docker volume rm nginx-log
    If, after powering on or running ip-change, you attempt to execute to the container: 
    [root@hid ~]# docker exec -it d00be2ea7a01 /bin/bash
    and the result is: 
    OCI runtime exec failed: exec failed: container_linux.go:296: starting container process caused "process_linux.go:78: starting setns process caused \"fork/exec /proc/self/exe: no such file or directory\"": unknown
    For Docker, run the following to restart the service: 
    [root@hid ~]# service docker restart
Redirecting to /bin/systemctl restart docker.service

    For Podman, run the following to restart the service:

    [root@hid ~]# systemctl restart rainier
  • If the Ops Center Administrator installation succeeded, but registering with Ops Center Common Services failed, run the setupcommonservice command after the installed Ops Center Administrator goes online.

Required

  • Log on to Ops Center Administrator to verify the installation.
  • Generate and install a signed SSL certificate. By default, the Ops Center Administrator installation package comes with a self-signed certificate that you can use to initially log in to Ops Center Administrator.

Optional