Unsupported Screen Size: The viewport size is too small for the theme to render properly.

Webhosting Data Replication?

  • This topic is empty.
Viewing 2 reply threads
  • Author
    Posts
    • #8868
      mv-it.com
      Member

      Here is my scenario….

      I currently resell hosting and my shared reseller server has had about 95% uptime (down timeis always percieved longer than uptime). A week ago the hard drive in the shared server crashed, the recovery process is still taking place. As a result I am looking into a failover system. Here's the plan… Please tell me if SJA would work in this (hairbrained) idea:

      Configure 2 reseller hosting accounts with 2 different data centers.

      Use a 3rd party DNS to manage failovers (60 sec ttl on DNS)

      Synchronize files using syncback to my computer and back to the servers.

      Synchronize MySQL with SJA (Enterprise?) to my computer and back to the servers.

      I will not have root access to the servers. I will have a similar access to the databases.

      Now here come the questions…

      What kind of setup would be required at the servers?

      Would I need any special access to the server?

      What kind of bandwidth can I expect to use in this process (sync 2 times a day/ 300k databases)? Would it read the entire database across the net?

      Can I implement this as a fire & forget? Set it up and walkaway and it will syncronize the 3 databases?

      Thank you

      Larry Wickham

    • #17223
      Ritesh
      Member

      You probably want to use SQLyog Data Synchronization Tool.

      Its available from Tools -> PowerTools -> Database Synchronization Wizard

    • #17224
      Shadow
      Member

      No special setup is needed, just make sure that database structures are the same on all the computers and use the same MySql version. You may use SQLyog's data structure sync tool to ensure the first criterion.

      You'll need SELECT, INSERT, UPDATE, DELETE privileges to all tables to be synced.

      300k databases are small, even a dial-up connection would be sufficient, although a simple DSL or leased line would be a more reliable method. Whole databases will not be copied (apart from the first sync), SJA uses concat_ws() to generate a checksum for each record to be compared.

      SJA is scheduled by OS' built-in scheduler tool, once you are convinced that the sync does what you had expected, then you can leave the job alone.

Viewing 2 reply threads
  • You must be logged in to reply to this topic.