MySQL Lists are EOL. Please join:

List:General Discussion« Previous MessageNext Message »
From:Sid Lane Date:September 15 2008 1:14pm
Subject:Slave SQL thread stopping w/o throwing error?
View as plain text  
hey,

I have three servers in a ring that have never had a problem before this
(well, at least none whose cause wasn't immediately obvious/easily fixed)
but starting at ~4:30 EDT yesterday the slave sql thread has started
spontaneously dying on different nodes for no apparent reason.

here is the excerpt from the error log of the 1st occurance (they're all the
same, just different timestamps @ log pointers):

080914  4:30:17 [Note] Slave SQL thread exiting, replication stopped in log
'mysqld.010810' at position 36177751

and the show slave status didn't report a failed statement (again,
last_errno always 0 & last_error always null):

*************************** 1. row ***************************
             Slave_IO_State: Waiting for master to send event
                Master_Host: <sanitized for public viewing>
                Master_User: <sanitized for public viewing>
                Master_Port: 3306
              Connect_Retry: 60
            Master_Log_File: mysqld.010810
        Read_Master_Log_Pos: 74094010
             Relay_Log_File: mysqld-relay-bin.021456
              Relay_Log_Pos: 931724
      Relay_Master_Log_File: mysqld.010810
           Slave_IO_Running: Yes
          Slave_SQL_Running: No
            Replicate_Do_DB:
        Replicate_Ignore_DB: mysql,gumsdata,mysql,gumsdata
         Replicate_Do_Table:
     Replicate_Ignore_Table:
    Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
                 Last_Errno: 0
                 Last_Error:
               Skip_Counter: 0
        Exec_Master_Log_Pos: 36177751
            Relay_Log_Space: 11129249
            Until_Condition: None
             Until_Log_File:
              Until_Log_Pos: 0
         Master_SSL_Allowed: No
         Master_SSL_CA_File:
         Master_SSL_CA_Path:
            Master_SSL_Cert:
          Master_SSL_Cipher:
             Master_SSL_Key:
      Seconds_Behind_Master: NULL

has anyone ever seen this behaviour and/or have any idea how to go about
diagnosing?  I've written a little perl nanny that loops running a show
slave status and restarts the sql thread if it stops which seems to send it
on it merry way (till the next time) but I'd rather solve the problem than
mask it.

please help!

Thread
Slave SQL thread stopping w/o throwing error?Sid Lane15 Sep
  • Re: Slave SQL thread stopping w/o throwing error?Sid Lane15 Sep
    • Re: Slave SQL thread stopping w/o throwing error?Ananda Kumar16 Sep