CMON 1.4.1 released - upgrade instructions

Follow

Comments

47 comments

  • Avatar
    David Majchrzak

    Nevermind, I pressed Cluster Registrations link on top and then Cluster and it came back.. 

    0
    Comment actions Permalink
  • Avatar
    Aigars Polis

    Hi,

     

    Command to add repository for Debian is not correct ! When I execute I get only OK in "/etc/apt/sources.list.d/s9s-repo.list".

    ##########

    cat /etc/apt/sources.list.d/s9s-repo.list

    OK

    ############

    Can you please provide correct command how to add repo for Debian?

    0
    Comment actions Permalink
  • Avatar
    Aigars Polis

    I also got errors when importing data in DB.

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_db_mods-1.2.8-1.2.9.sql

    Enter password:

    ERROR 1054 (42S22) at line 5: Unknown column 'component' in 'simple_alarm'

     

    mysql -uroot -p < /usr/share/cmon/cmon_data.sql

    Enter password:

    ERROR 1054 (42S22) at line 14: Unknown column 'cid' in 'field list'

    0
    Comment actions Permalink
  • Avatar
    Aigars Polis

    Correct script for Debian should be:

                                              

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_db.sql

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_db_mods-1.2.6-1.2.7.sql

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_db_mods-1.2.7-1.2.8.sql

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_db_mods-1.2.8-1.2.9.sql

    mysql -uroot -p -f cmon < /usr/share/cmon/cmon_data.sql

    0
    Comment actions Permalink
  • Avatar
    Johan

    Hi Aigars,

    We will look into it, particularly about the repo setup, but the sql scripts seems already to be written in the correct order.

    Best regards

    Johan

    0
    Comment actions Permalink
  • Avatar
    Joachim Heidenreich

    Works like a charm on Debian 7 

    0
    Comment actions Permalink
  • Avatar
    Tobi Schulz

    Just upgraded from 1.2.8 to 1.2.9 and now I am stuck in redirect loops.

    Deleted all cookies from the domain, even restarted the box just in case - still redirects forever after login.

    We are using nginx on CentOS following exactly the instructions given here:

    http://www.severalnines.com/blog/how-configure-clustercontrol-run-nginx

    Have there been changes to the file/url structure? (Guess so...)

    TIA!!

    0
    Comment actions Permalink
  • Avatar
    Tobi Schulz

    Just for clarification, this is what the nginx access log gives:

    62.xxx.xx.xxx - - [18/Mar/2015:22:32:18 +0100] "GET /clustercontrol/ HTTP/1.1" 302 5 "-" "Mozilla/5.0"

    62.xxx.xx.xxx - - [18/Mar/2015:22:32:18 +0100] "GET /clustercontrol/users/login HTTP/1.1" 302 5 "-" "Mozilla/5.0"

    62.xxx.xx.xxx - - [18/Mar/2015:22:32:19 +0100] "GET /clustercontrol/ HTTP/1.1" 302 5 "-" "Mozilla/5.0"

    62.xxx.xx.xxx - - [18/Mar/2015:22:32:19 +0100] "GET /clustercontrol/users/login HTTP/1.1" 302 5 "-" "Mozilla/5.0"

    and so on and so on.

    Seems like the login mechanism has changed or something along those lines.

    0
    Comment actions Permalink
  • Avatar
    Krzysztof Ksiazek

    Hi Tobi,

    Please add following rewrite rules to the server section of your nginx config file. In 1.2.9 couple of .htaccess files were added and nginx doesn't support them.

    location / {

    rewrite ^/$ /app/webroot/ break;

    rewrite ^(.*)$ /app/webroot/$1 break;

    }

    location /app/webroot/ {

    if (!-e $request_filename){

    rewrite ^(.*)$ /index.php?/$1 break;

    }

    }

    0
    Comment actions Permalink
  • Avatar
    Tobi555

    You're a star!

    That works fine now!

     

    0
    Comment actions Permalink
  • Avatar
    Damodar Chettani

    Hi Johan,

    I have 1.2.9.657 Cluster control. When i executed the procedure I found a few error:

    Cannot open: clustercontrol-controller-1.2.10-727-x86_64.rpm. Skipping.

    When I checked my actual version I see: 

    |
    ClusterControl UI version:
    | 1.2.10.5dbcb85
    |
    |
    ClusterControl CMON Version: 
    | 1.2.9.713
    |
    |
    CMON API Version:
    | 1.2.10.4103a2a
    |

    It is correct?

    0
    Comment actions Permalink
  • Avatar
    Petr Hrabal

    Hi,

    in Debian manual section there is a typo:

    wget -O cmon-controller-1.2.10-727-x86_64.rpm http://www.severalnines.com/downloads/cmon/cmon-controller-1.2.10-727-x86_64.rpm

    Ofc it should be .deb

    also did someone have problem with those error? (using wheezy update point 7.8)

    dpkg: unrecoverable critical error, aborting:

    File List package "netcat-traditional" does not contain the final newline

    dpkg: unrecoverable critical error, aborting:

    File List package "net-tools" does not contain the final newline

    Petr

     

     

    0
    Comment actions Permalink
  • Avatar
    Aigars

    Hi,

     

    After upgrade I got message "cluster is not running. Please, verify RPC_TOKEN settings". Where I should look for this setting?

    0
    Comment actions Permalink
  • Avatar
    Alex

    Hi

    How did you perform the upgrade? Try logout and then login again from the web application. It should have automatically updated the clusters with the correct RPC keys.

    If that still doesn't work then

    • Check that your <webroot>/clustercontrol/bootstrap.php contains a parameter called RPC_TOKEN

    • If you only have a single /etc/cmon.cnf file then check that you have a parameter called rpc_key with the same value as the above RPC_TOKEN

    • If you have /etc/cmon.d/cmon_N.cnf file(s) then check that those files have a rpc_key parameter that matches below output

    mysql -ucmon -p -e "select id as clusterid, token from dcps.clusters"

     

     Best Regards,

    Alex

    1
    Comment actions Permalink
  • Avatar
    Aigars

    Hi,

     

    Updated was performed with APT.

    Looks like my original cmon.cnf file was replaced in upgrade procedure. When I restored old one then ClusterController started up without errors.

    Thanks.

    0
    Comment actions Permalink
  • Avatar
    Joachim Heidenreich

    Hi,

    I had the same problem updating to cmon 1.3.0 with APT on Ubuntu. Performing the steps Alex posted fixed the error.

    BR

    0
    Comment actions Permalink
  • Avatar
    Petr Hrabal

    Hi, Upgraded to latest version and hit a strange glitch 

    my galera cluster is graphed normaly but its reported as failed.

    colums Status,wsrep cluster size,local quee, flowcontrol and cert deps distance seems to be empty/undefined...

    In reality cluster size is 3 and all nodes are in primary state. Please note that I was upgrading galera as well and I am now on lastest percona version for 5.6

    Please advise where/how to check what is wrong.

    BR petr

     

    0
    Comment actions Permalink

Please sign in to leave a comment.

Powered by Zendesk