Layer7 API Management

 View Only
  • 1.  Upgrade fails from 8.4 to 9.0

    Posted Nov 12, 2019 12:05 PM
    ​Running 8.4 and trying to  upgrade to "Layer7_PlatformUpdate_64bit_v9.0.00.L7P" as the first step to 9.4 and get this error:

    Please wait while the patch is installed ...
    Patch ID Layer7_PlatformUpdate_64bit_v9.0.00 (Upgrades the Layer 7 64bit Platform to the 9.0.00 version. This patch requires that the 8.4.00 Platform Update be installed first. It also requires a restart of the Layer 7 Appliance.) is ERROR, last modified on 2019-11-12T07:57:05-0800
    There were errors during the patch operation.


    It is on 8.4 and the status file is in place in the patches directory. 

    Any ideas?  Or where can we look up details of the failure?


    ------------------------------
    TBC
    ------------------------------


  • 2.  RE: Upgrade fails from 8.4 to 9.0
    Best Answer

    Posted Nov 13, 2019 03:29 AM

    Hi,

    Usually it's either a disk space issue (in /tmp, /opt or /home), or a permissions issue. Not much else (since 8.3 version) causes any issues.

    Please check that there's at least 1GB in all those mount points and that the L7P file is owned by gateway (not ssgconfig!) and with 666 permissions.

    I use the following guide for all my patches and upgrades. (I have upgrades/patched at least 500-1000 nodes over the last 6 years and usually performed it on to up to 30 at a time using iTerm2 broadcast, so I'm quite confident)

    First take the node(s) affected out of the load balance pool)

    df -h /opt /tmp /home

    rm -f ~ssgconfig/*.L7P

    rm -f /tmp/*.L7P

    rm -f /opt/SecureSpan/Controller/var/patches/*.L7P

     

    Download L7P, do permissions and move it to ssgconfig dir:

     

    cd /tmp

    Up/Download L7P file (Example used below)

    chmod 666 CA_API_PlatformUpdate_64bit_v9.X-RHEL-2018-10-16.L7P

    chown gateway.gateway CA_API_PlatformUpdate_64bit_v9.X-RHEL-2018-10-16.L7P

    mv CA_API_PlatformUpdate_64bit_v9.X-RHEL-2018-10-16.L7P ~ssgconfig/

    su – ssgconfig

     

    Do upload and patch (using ssgconfig menu or command line patch.sh)

     

    service mysql restart on all

    mysql_upgrade on all (This is sometimes included in the L7P process, but no harm will be done by trying multiple times)

    (If you patched a full cluster in parallel)

       /usr/bin/reboot
    on MASTER

         Wait for Master to come up, then /usr/bin/reboot on all other nodes in cluster

    Otherwise just reboot the node in question.

    If you have a failed installation you will also need to remove the .status file from the patch directory and ensure the L7P file is in ~ssgconfig with the correct permissions again (it gets owned by ssgconfig during the process)

    Regards

    Vince



    ------------------------------
    Senior Architect
    Apiida AG
    Https://www.apiida.com
    ------------------------------



  • 3.  RE: Upgrade fails from 8.4 to 9.0

    Posted Nov 13, 2019 07:02 AM
    Hello TBC,

    Kindly Check Gateway /opt, /tmp size always before starting the upgrade,I am suspecting disk space issue.

    /opt available space: 39% of /opt folder
    /tmp available space: 4% of /tmp folder