No announcement yet.

Problem with RBR replication, HA_ERR_END_OF_FILE

  • Filter
  • Time
  • Show
Clear All
new posts

  • Problem with RBR replication, HA_ERR_END_OF_FILE

    After preparing a slave server I have an error after a while running:
    Could not execute Update_rows event on table tbl; Can't find record in 'tbl', Error_code: 1032; handler error HA_ERR_END_OF_FILE; the event's master log mysql-bin.72142, end_log_pos 6387298

    I don't know what could be the problem. This master-slave configuration have been running for years without any problem in SBR mode and now is getting trouble changing to RBR.

    How to debug it ?

  • #2
    Your replica has different data than the master does. pt-table-checksum and pt-table-sync are your friends.


    • #3
      Yes, but why I have different data ? Someone is modifying data on slave server ? Is strange because is only for read, and replication is configurated with RBR mode fron scratch...

      I will configure slave server with read only mode to know if some of my developers is using it to modifying data ...


      • #4
        Are you absolutely sure that you got your master and slave in sync when you started the replication?
        Since if they aren't completely in sync then everything will look like it works for a while and data will be replicated to the slave. Until one day an UPDATE comes along and wants to update one of the records that went missing on the slave in the "gap" of the replication.


        • #5
          Read "High Performance MySQL." There is an extensive section about why this problem can happen and how to detect and fix it.


          • #6
            I am completely sure because I am taking a backup with xtrabackup and then I have configured rreplication with last binlog info of xtrabackup log.

            Where is "High Performance MySQL"?


            • #7
              Ok, seems that I know the problem now, xtrabackup is not getting a correct .par file, all of them are 0 size files... big problem with backups!


              • #8
                Please report that as a bug IF you are using the latest version of XtraBackup. I think a bug related to that problem was fixed a while ago. Of course, this could be a new bug.


                • #9
                  I will investigate further. I am using 1.6.3 version. Could be out of space of destination backup directory ?


                  • #10
                    It's possible -- it could also just be a bug


                    • #11
                      This was not a bug, was a problem on the destination backup directory, it was fulll but it seems like xtrabackup is not sending any error while copying files so I don't detect anything until now.

                      I need to update xtrabackup to have a correct redirect of errors from app. Now I have tried to update but I get Segmentation fault on a Debian... I will ask in correct forum space.

                      Thank you guys!