CMON 1.4.1 released - upgrade instructions

Follow

Comments

47 comments

  • Avatar
    Johan

    Thanks,

    We will also update the instructions shortly to accomodate for upgrade to 1.1.36.

    Best regards,

    Johan

    0
    Comment actions Permalink
  • Avatar
    Sistemas Cloud

    Hello,

    I'm on 1.1.36a version. Are there any db_mods to  upgrade from this version??

    Thanks,

    0
    Comment actions Permalink
  • Avatar
    Johan

    Hello,

    Yes, this one:

    cmon_db_mods-1.1.36-1.2.0.sql 

    Best regards,

    Johan

    0
    Comment actions Permalink
  • Avatar
    Marc Bernard

    The file does not exist: http://www.severalnines.com/downloads/cmon/cmon_db_mods-1.1.36-1.2.0.sql

    Where would it be instead ?

    0
    Comment actions Permalink
  • Avatar
    Johan

    Hi Marc,

    It is included in the cmon-controller RPM or the cmon tar ball:

    /usr/share/cmon/cmon_db_mods-1.1.36-1.2.0.sql 

    or for the tar ball (when unpacked and installed in std location):

    /usr/local/cmon/sql/cmon_db_mods-1.1.36-1.2.0.sql 

    0
    Comment actions Permalink
  • Avatar
    Marc Bernard

    I have followed instructions above, I confirm that cmon_db_mods-1.1.36-1.2.0.sql is not in /usr/share/cmon/ or /usr/local/cmon/sql/ after verifying, I found out your post has an error; You are missing the R in controller. Here is the correct syntax.

    rpm -Uvh cmon-controller-1.2.0-1.x86_64.rpm

    0
    Comment actions Permalink
  • Avatar
    Ethan Hall

    Johan,

    We're still running 1.1.14.  How many of the intermediate DB mods do I need to run to bring us up to current?

    Thanks!

    Ethan

    0
    Comment actions Permalink
  • Avatar
    Graham Green

    There are is a minor error with your schema upgrade commands, there should be no space between the -p and the password, for example the first command should read:

    mysql -f -h127.0.0.1 -u cmon -pcmon < /usr/share/cmon/cmon_db.sql

    Additionally I'm having problems adding an existing cluster; There are a number of issues:

    1) There is no documentation on how to add an existing cluster

    2) Using an IP address for the Cluster Control Node results in an Alert "Invalid 'ClusterControl Controller Hostname/IP'. Please use different IP address for the Controller" Changing this to a hostname (to which that IP address belongs) makes the error go away.

    3) It's not clear what I should be providing for the Select SSH Key field. I already established and SSH key pair when I did the initial cluster control configuration for the first cluster and I provide the path to that in the SSH Key Path field. I guess I don't understand the need to add a keypair when I want to use already established credentials. If I try uploading the private key used by the root user I get a message saying "Please wait... Saving..." and it just sits there.

    0
    Comment actions Permalink
  • Avatar
    Johan

    Graham,

    The line should read  "mysql -f -h127.0.0.1 -u cmon -p      cmon < /usr/share/cmon/cmon_db.sql"   .

    -p is there for the user to enter the password when prompted. The subsequent 'cmon' is not the password but the cmon database.

    Regarding 1) we have done:

    http://support.severalnines.com/entries/45804993-Add-Existing-Cluster  and we have simplified the "adding an existing cluster" a lot. This simplification render 2) and 3) obsolete. Please give it a try.

    BR

    johan

     

    0
    Comment actions Permalink
  • Avatar
    Graham Green

    Interesting re the password, I got failures until I removed the space (after looking at the.sql file and seeing it had a use statement for the database). Perhaps I was typing the password incorrectly.

    Since posting my initial post I have a new problem. The UI was working fine for me after the upgrade, however I closed the browser tab and I'm now presented with an error: "Table ldap_group_roles for model LdapGroupRole was not found in datasource default." when I try to access the GUI.

    I have checked for the existence of that table and found in the database dcps which I believe was created as part of the upqrade. Any clues as to what might be happening here? I have not tried to configure the system to use LDAP.

    0
    Comment actions Permalink
  • Avatar
    Alex

    Hi

    Can you try this:

    1) Logout the CC application

    2) rm -f <your www root dir>/clustercontrol/app/tmp/cache/models/*

    3) Clear the browser cache and login again

    Did that work?

    Best Regards,

    Alex

    0
    Comment actions Permalink
  • Avatar
    Graham Green

    I am no longer getting the error but the GUI is not listing my existing cluster (see attached screenshot)

    0
    Comment actions Permalink
  • Avatar
    Alex

    Hi

    Can you provide the content of the browser debug console? For example in Chrome you have to go to View->Developer->Javascript Console.

    Also these logs

    • <your www root>clustercontrol/app/tmp/logs/error.log
    • <your www root>/clustercontrol/app/tmp/logs/debug.log

     

    Thanks.

     

    Best Regards,

    Alex

    0
    Comment actions Permalink
  • Avatar
    Graham Green

    Here is the output of the developer console in Chrome

    Uncaught TypeError: Cannot read property 'length' of null production.js?t=1398270342:31

    Failed to load resource: the server responded with a status of 404 (Not Found) http://galeracc01/clustercontrol/js/ClusterControl.js?_dc=1398295139197

    Uncaught TypeError: object is not a function (index):3

    event.returnValue is deprecated. Please use the standard event.preventDefault() instead.

    The error log is quite large (I have included the tailed entries after I tried loading the page again) if you need more let me know.

    0
    Comment actions Permalink
  • Avatar
    Graham Green

    The tar file for the cc-ui 1.2.6 upgrade doesn't contain ClusterControl.js. That file also does not exist on another install of CC which is running CC 1.2.5.

    0
    Comment actions Permalink
  • Avatar
    Alex

    Thanks for the logs. The web team is still offline and we'll update you as soon as they have some time to look into this.

    The ClusterControl.js is compiled into the production.js so that's OK. I haven't seen that error message is showing up in the console before though.

    What version of Chrome are you using?

    Thanks.

    Best Regards,

    Alex

    0
    Comment actions Permalink
  • Avatar
    Alex

    I was able to re-produce this. It seems to be a browser cache issue.

    Can you try this:

    1) Logout of the CC application

    2) Press shift+reload on the login page then login.

    3) You should see a 'search' button in the new version. Is there? If not try shift+reload on the page.

    4) If you get the same issue try 1-3) again.

     

    Did that work for you?

     

    Best Regards,

    Alex

     

    0
    Comment actions Permalink

Please sign in to leave a comment.

Powered by Zendesk