Monitoring High Availability on Galera Clusters for MySQL using MONyog

  • Chetan
  • July 29th, 2015
  • MONyog, MySQL
  • Monitoring-High-Availability-on-Galera-Cluster-for-MySQL-using-MONyog

    For every online business, there are millions of transactions happening every second including crucial revenue bearing transactions. Therefore downtime, both planned and unplanned has a cost attached to it and results in lost revenue every second. As the technology improved to meet this demand, the importance of High Availability (HA) became the focus point.

    The introduction of Galera Cluster for MySQL bought true Multimaster Cluster thanks to its synchronous replication. In this article, we’ll show how using MONyog in tandem with Galera provides a simple and best-in-class High Availability solution for MySQL users.

    How to register a Galera Cluster Node in MONyog

    1. Enter the Hostname, port and credentials in “MySQL settings”, test the connection and save it.
    2. Enable “Galera” group from “Customize” > “Manage monitor groups”
    3. In Monitors page, select the “Galera” group and you’re all set.


    Setting up Galera Charts

    With the monitors setup, it would be nice to have a chart to monitor the health of the MySQL cluster. In the Charts tab, click on the gear icon on the right and select ‘Add a new Chart’.

    Let’s create a Chart to view bottleneck in replication between the Galera nodes. This can be done by monitoring the status variable wsrep_local_send_queue_avg
    Add a Series Caption and in Series Value add: [MONyog.MySQL.GlobalStatus.wsrep_local_send_queue_avg] also set the ChartValue to Current.


    Let’s create another Chart to show the %time Galera replication was paused.
    Add a Series Caption and in Series Value, add: [ToInt(MONyog.MySQL.GlobalStatus.wsrep_flow_control_paused) * 100]

    Once the Galera Charts are setup, your Monitors tab will look something like this.


    Galera monitors shipped with MONyog

    Here are the monitors currently available in MONyog. You can hover on them to view advisors and customize as per your requirement.
    Galera monitors shipped with MONyog

    Alerts for Galera

    With Galera Charts in place, all you need is alerts to complete the Galera Cluster monitoring setup. Using this, you can make MONyog alert you when a condition is triggered.

    MONyog supports one default alert for wsrep_ready status variable in the Galera Monitors tab. You can manually enable alerts for any variable. For example, if you want to setup an alert to find out network bottlenecks, here’s how you do it:

    1. Go to Monitors tab, select the Galera group, click on ‘Find network bottleneck’.
    2. In the popup, click on Customize.
    3. On the top-right part of the page you’ll find an option to ‘View advanced’. Click on it and scroll down to see a textarea for ‘Alert condition’.
    4. Now you can write your own alert condition like this:
      function() {
      if( parseInt(this.Value) >= 1)
      return "Critical";
      else if ( parseInt(this.Value) > 0.5)
      return "Warning";
      else return

    If you’re new to MONyog, give it a try and start monitoring your MySQL/MariaDB Galera Clusters. If you have any feedback or questions, drop a comment below.

    Introducing actionable charts in MONyog MySQL Monitor

  • Chetan
  • July 23rd, 2015
  • MONyog, MySQL
  • While MONyog MySQL Monitor is being embraced by DBAs and DevOps around the world, we are constantly adding more features to make MONyog the most complete MySQL monitoring tool.

    MONyog ships 40+ MySQL and system metric charts. It also lets you add custom charts. With the latest update, we have made Charts more powerful by adding few important features to help you find problem queries with ease.

    Here’s a 1 minute video to walk you through the new features.

    Existing customers can download MONyog 6.33 from Customer Area. To evaluate MONyog 6.33, please download a 14-day trial.

    We are really excited about this update and we hope you like the new features. Please feel free to share your feedback.

    Team MONyog

    MONyog MySQL Monitor 6.33 Has Been Released

  • peter_laursen
  • July 23rd, 2015
  • MONyog, Releases
  • Important notice: With the previous 6.32 release we changed registration keys. Please have your new key available when upgrading from a pre-6.32-version. As always you will get it from our Customer Portal.

    Changes (as compared to 6.32) include:


    * Added monitoring of the MySQL –tmpdir volume with the “Free Disk Space” Monitor for Linux (introduced in 6.32).
    * Added an option to view X-Axis labels for large time based charts.
    * The Performance_Schema -based “query sniffer” has been optimized.

    Bug Fixes:

    * Fixed a rare case, where “Charts” page was not loading properly.


    MONyog MySQL Monitor 6.32 Has Been Released

  • peter_laursen
  • July 16th, 2015
  • MONyog, Releases
  • Important notice: With this release we have changed registration keys. Please have your  new key available when upgrading. As always you will get it from our Customer Portal.

    Changes (as compared to 6.31) include:


    * Added a “Free Disk Space” Monitor for Linux. This new feature will collect information directly from the file system (unlike the existing ‘Disk Info’ that uses MySQL “SHOW TABLE STATUS”). Information is collected from volumes storing both MySQL data directory and InnoDB data directory (if different), and user will be alerted if either crosses a specified threshold.
    * In Charts page added an option to select a time interval for display by zooming a chart using mouse and similar input devices (touchpad etc.). Note that zooming one chart will change the timeframe of all the charts displayed.
    * Added an option to view queries for a selected time range in time range based charts in Charts page.

    Bug Fixes:

    * In rare cases, Monitors page could show wrong data for Linux counters if the HISTORY timeframe was selected for the page.  Other available timeframes were not affected.
    * When monitoring a MariaDB server using a galera-based replication setup we will now execute SHOW ALL SLAVES STATUS rather than just SHOW SLAVE STATUS. Before this MONyog replication page showed an empty result and alerts were not sent. However due to limitations with current MONyog architecture we are only able to handle the first row of SHOW ALL SLAVES STATUS with this release, if more are returned, and only information from one slave node will be available.


    MONyog MySQL Monitor 6.31 Has Been Released

  • peter_laursen
  • June 25th, 2015
  • MONyog, Releases
  • Changes (as compared to 6.3) include:

    * What is plotted in “Charts” (formerly “Dashboard”) page may now be specified as a time interval and not only as a number of latest sample points. This increases usability of the page.
    * Long running query log now includes action (kill/notify/notify and kill) and the time of action.
    * Locked and locking queries tab in Real-Time interface is optimized and loads faster.
    * Added w    ildcard filtering using the “*” character as wildcard for user and host in Query Analyzer for Slow Query Log.

    Bug Fixes:
    * MONyog failed to send notification for an InnoDB deadlock if deadlock details (as recorded by SHOW ENGINE INNODB STATUS) contained double quotes.


    MONyog MySQL Monitor 6.3 Has Been Released

  • peter_laursen
  • May 20th, 2015
  • MONyog, Releases
  • Changes (as compared to 6.27) include:

    * a log of long running queries can now be generated and downloaded as CSV in Sniffer interface (applies only to Processlist mode of sniffer).
    * Real-Time is optimized and loading a session is much faster now. Further it shows only data from top 200 open tables and databases.
    * Wayback Machine now includes user and host information as well.
    * Queries in Wayback Machine can now be exported as CSV.
    * MONyog can now send long running query notification to specific selected users only.
    * Wildcard filtering using the “*” character as wildcard for user and host is introduced in Sniffer (applies only to Processlist mode of sniffer).
    * Same wildcard filtering implementation is also introduced for Ignore user and host in Long running query option in Sniffer (applies only to Processlist mode of sniffer).
    * MONyog now logs which user killed a thread in Processlist tab.
    * SQLite library bundled in MONyog is now upgraded to the latest version. This results in performance improvement for sniffer in particular where it can be up to ~50%.

    Bug Fixes:
    * In some cases, MONyog logged SQLite errors when CSOs (Custom SQL Objects) were enabled even when the CSO worked as expected
    * Small UI fixes.


    SQLyog MySQL GUI 12.12 Released

  • peter_laursen
  • May 19th, 2015
  • Releases, SQLyog
  • Changes as compared to 12.11 include: 

    * XML files can now be imported from GUI controls in table menu/context menu provided that details of the XML follows one of the three supported XML-schemes supported by MySQL LOAD DATA XML syntax. This includes XML files generated with SQLyog and mysqldump.

    Bug fixes:
    * Server name in object browser was displaying wrong in some cases.
    * JSON export added in version 12.11 could add an extra delimiter what could cause an error in the environment where the JSON was used.
    * Profiler tab was getting focus instead of Messages tab in some cases if query was executed by placing cursor inside query and if more queries appeared in the Query tab and queries were executed ‘one by one’ (it did not happen when highlighting the query).
    * In ‘Export as’ dialog, SQLyog did not remember the escape character used if escape character was defined as “\\”. It would then replace “\\” with “\”.
    * In some cases SQLyog wasn’t displaying warnings for the query executed. Most common case was when a function was used with an argument of incorrect type causing server-side truncation of the argument (example: “SELECT ABS(‘a’);”)

    * Increased the dialog and column sizes for index and foreign key popup dialogs for better usability.


    Life of a DBA in GIFs

  • Chetan
  • May 11th, 2015
  • MySQL
  • A Database Administrator experiences a wide range of emotions. It could be one those endless meetings, friendly disagreements with fellow developers, getting something approved from managers or preparing your junior DBAs for bigger battles. Each day is a challenging one. We’ve tried to compile a list of GIFs which every DBA will be able to relate to.

    5 minutes before deployment

    Writing the most epic answer Stack Exchange has ever seen and press F5 to “Submit” and end up refreshing the page

    When a DBA.StackExchange answer gets 500+ upvotes!

    Slightly “re-factoring” the developers’ code after a code review

    Training the junior DBA

    The junior DBA trying to figure out the production cluster

    When you unknowingly fix the client’s problem

    When the project manager starts questioning my work estimates

    When asked why I’m allowed to query all the databases?

    When a developer says he made few changes to the database

    When I shut off the marketing team’s access to run reports on the production server

    PS: Don’t forget to visit DBA Reactions for a daily dose of GIFs chronicling the highs and lows of a DBAs life.

    If you’re not a DBA, make sure you share it with one and watch them nod their head in approval. If you have anything to share, feel free to use the comments section below.

    Rockstar DBA: An Interview with Manojit Paul

  • Karthik Suroju
  • April 30th, 2015
  • MONyog, MySQL, Rockstar DBA
  • Rockstardba
    This article is a part of our Rockstar DBA initiative, a platform for thought leaders to share their knowledge and make a positive contribution towards the database management community.

    Manojit Paul, with more than 19 years of professional DBA experience in enterprise database solutions, especially Sybase product set and database infrastructure is our first Rockstar DBA.

    Let’s hear from Manojit about how he manages MySQL databases, why he loves MONyog, his advice to young DBA’s and more.

    How do you manage your MySQL databases?
    We presently have 9 MySQL servers in production trading environment, all being managed via a host of automated scripts which I’ve created and manual intervention when issues are reported. We deploy MONyog for managing 5 of these critical MySQL instances.

    What are your biggest challenges when dealing with MySQL databases?
    MySQL is not real enterprise class (evident from Gartner Reports omissions) and the rest of our database infrastructure relies on Sybase ASE and replication platform used for actual trading and logistics. We do have selective schema replication in MySQL but lack of proper scaling and optimization is a challenge.

    How does MONyog help you? What’s your favorite feature?
    We deploy MONyog for the MySQL monitoring and its web based dashboard helps with quicker problem identification and resolution. What I really like about MONyog is the dashboard with real-time statistics along with the charts displaying the overall performance metrics.

    Do you use other tools to manage your MySQL servers?
    We do not use any other MySQL related tools except the dbadmin scripts I have written.

    According to you, what is the future of MySQL?
    I want MySQL to compete with Percona/MariaDB and keep up with the feature sets that are still missing, like object level replication and online database backups and more support for the community version too.

    What blogs/forums/magazines do you follow regularly?
    Gartner reports related to database technology
    ISUG (sybase)

    What advice can you offer to young enthusiasts who want to pursue a career as DBA?
    I personally follow these basic rules. Not sure who gave these to me, they are helping me all along my professional life.

    1. Assume nothing
    2. Have contingencies
    3. Simplify
    4. Plan and Execute with precision
    5. Strive for Zero Defect work

    We are thankful to Manojit Paul for sharing his thoughts with us. We would love to hear your feedback as comments about our Rockstar DBA initiative.

    Sharing is caring. If you are a DBA who has something significant to offer to the community, we welcome you to be our next Rockstar DBA. Just send an email to us at

    SQLyog MySQL GUI 12.11 Released

  • peter_laursen
  • April 27th, 2015
  • Releases, SQLyog
  • Important notes: .
    This release requires that user re-enters his registration key. Please have it available when upgrading. The key itself is the same as used by previous 12.x versions.
    * This release includes a new HTTP tunneler script. Functions deprecated in recent PHP versions are not used anymore and the script now also follows PHP strict standards. Users using HTTP tunnel must replace the script on their server(s).

    Other changes as compared to 12.1 include: 

    * Added option to export table data and result sets as JSON.

    Bug fixes:
    * On Windows systems using a recent wininet.dll, setting the HTTP timeout higher than 30 seconds had no effect and SQLyog would fire same query 3 times if HTTP tunneling was used and query executed on server for more than 30 seconds.
    * Fixed a performance regression with “execute SQL script”.

    The new icons and other graphics introduced in 12.1 now have slightly darker and brighter colors.  As a result the UI is now ‘less pastel’ than before.
    Changes in 12.1 to the ‘twilight’ theme (default before 12.1) have mostly been reverted due to user requests.
    * Visual Data Compare UI now has the same 12.1 -style overall look as the main program window etc. has.

    SQLite  was updated to version This will result in performance enhancement of up to 80% for Autocomplete with large local SQLite databases.


    Page 1 of 4612345...102030...Last »