Announcing: MySQL Fabric 1.5.5 GA is Available!

The MySQL Fabric team is pleased to announce that a new (GA) release is now available. This release includes several improvements in terms of stability and usability (bug fixes and feature enhancements). In a nutshell:

  • Two new types of users for backup and restore are now available allowing a better managements of the required privileges for each user type. More information:
  • An example script for pacemaker was added to the source packages to help provide an HA solution using a single Fabric node. More information:
  • The Connector/Python minimum requirement has changed. Fabric now requires versions >= 2.0.4 and < 2.1, or >= 2.1.2.
  • Added signal handlers for supported POSIX platforms as follows.
    • SIGUSR1 signal can be sent to the Fabric process to dump the stack trace of all threads to the current logger or STDERR.
    • SIGINT signals are caught by Fabric to avoid killing the process by mistake using <Control>+C. Note that the operation will abort after catching the signal three times.
  • The MySQL and XML RPC protocols were improved including the handling of MySQL connections (fix exceeded CPU usage) and threads (to avoid “Connection reset by peer” errors).
  • Improved support for 5.7.8 servers.
  • Fixed issues with executing some commands (e.g. server clone).
  • Improved input checking for address format “host:port” and error messages (e.g., when trying to build docs without Sphinx).

A complete list of all improvements can be found in our release notes:

Try it now and send us your feedback

MySQL Fabric release-1.5.5 GA is packaged together with MySQL Utilities and
is available for download from the following link.

MySQL developers website:

Fabric is also available on GitHub:

The documentation of MySQL Fabric can be found here:

Contributing Ideas:

About Paulo Jesus

Paulo Jesus is currently a Principal Software Developer at Oracle, and team lead of the MySQL Fabric (Core) team. He received the BEng degree in systems and informatics in 2001, and the MSc degree in mobile systems in 2007, both from the University of Minho (Portugal). He obtained his PhD degree in 2012, from the MAP-i doctoral program in computer science by the Universities of Minho, Aveiro and Porto (Portugal). He has many years of professional experience as a software developer in relevant companies, and taught informatics during 1 year. He is also an invited researcher at the HASLab / INESC TEC, University of Minho. His research interests include distributed algorithms, fault tolerance, and mobile systems.

4 thoughts on “Announcing: MySQL Fabric 1.5.5 GA is Available!

  1. How could we remove MySQL Fabric after “./ install” in case of we would like to downgrade or upgrade MySQL Fabric?

    1. Like many python applications, the way to remove a previous version and install a new one often requires a manual uninstall. This is the case for Fabric. Indeed, the steps that need to be followed are version specific.

      There have been several evolutions of Fabric and that is the reason why
      we have no standard uninstall/upgrade at this time.

      If you need help performing this operation, please contact us through
      one of the following communication channels:
      – For customers: open a support ticket at
      – For community users: open a feature request or bug at or post your problem at the Fabric Forum

      We will be happy to help you.

      Thank you for your interest in Fabric!

  2. The insert speed into shards with Connector/J is very slow according to the information in fabric.log table in the backing store.

    It seems that the connector spend a lot of time to get the shard information every time it connects.
    Besides, if there are only two shards with HASH mode, the two shards are pretty unbalanced (1:4 ratio in my case).

    Do both issue caused by mis-configuration or they are known issues?

    1. The use of a connection pool for Connector/J should help avoid this issue. Please confirm if you are using the latest Connector/J version and taking advantage of a connection pool. If the issue persist please open at against Connector/J.

      Regarding the shard balance issue, we are already aware of it. It is a known issue.

      If you need more detailed information for your specific scenario, please use the communication channels that I mention in the previous reply.

      I hope this helps.

      Thank you!

Leave a Reply

Your email address will not be published. Required fields are marked *

Please enter. * Time limit is exhausted. Please reload CAPTCHA.