database/mariadb
1
0
Fork 0
mirror of https://github.com/MariaDB/server.git synced 2025-07-30 16:24:05 +03:00
Code Activity

Bug#11747416 : 32228 A disk full makes binary log corrupt

Browse Source 
Binary log of master can get a partially logged event if the server
runs out of disk space and, while waiting for some space to be freed,
is shut down (or crashes). If the server is not stopped, it will just
wait endlessly for space to be freed, thus no partial event anomaly
occurs.  The restarted master server has had a dubious policy to send
the incomplete event to slave which it apparently can't handle.
Although an error was printed out the fact of sending with unclear
error message is a source of confusion.
Actually the problem of presence an incomplete event in the binary log
was already fixed by WL 5493 (which was merged to our current trunk
branch, major version 5.6). The fix makes the server truncate the
binary log on server restart and recovery.

However 5.5 master can't do that. So the current issue is a problem of
sending incomplete events to the slave by 5.5 master.

It is fixed in this patch by changing the policy so that only complete
events are pushed by the dump thread to the IO thread. In addition,
the error text that master sends to the slave when an incomplete event
is found, now states that incomplete event may have been caused by an
out-of-disk space situation and provides coordinates of
the first and the last event bytes read.
This commit is contained in:
Andrei Elkin
2011-09-29 14:14:43 +03:00
parent b140784fbc
commit b426043b7c
8 changed files with 110 additions and 27 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
sql/sql_repl.cc
View File

@ -352,7 +352,7 @@ Increase max_allowed_packet on master";
*errmsg = "memory allocation failed reading log event";
break;
case LOG_READ_TRUNC:
*errmsg = "binlog truncated in the middle of event";
*errmsg = "binlog truncated in the middle of event; consider out of disk space on master";
break;
default:
*errmsg = "unknown error reading log event on the master";
@ -447,6 +447,9 @@ void mysql_binlog_send(THD* thd, char* log_ident, my_off_t pos,
String* packet = &thd->packet;
int error;
const char *errmsg = "Unknown error";
const char *fmt= "%s; the last event was read from %s at %s, the last byte read was read from %s at %s.";
char llbuff1[22], llbuff2[22];
char error_text[MAX_SLAVE_ERRMSG]; // to be send to slave via my_message()
NET* net = &thd->net;
mysql_mutex_t *log_lock;
mysql_cond_t *log_cond;
@ -677,10 +680,8 @@ impossible position";
if (reset_transmit_packet(thd, flags, &ev_offset, &errmsg))
goto err;
my_off_t prev_pos= pos;
while (!(error = Log_event::read_log_event(&log, packet, log_lock)))
while (!(error= Log_event::read_log_event(&log, packet, log_lock)))
{
prev_pos= my_b_tell(&log);
#ifndef DBUG_OFF
if (max_binlog_dump_events && !left_events--)
{
@ -766,17 +767,6 @@ impossible position";
goto err;
}
/*
here we were reading binlog that was not closed properly (as a result
of a crash ?). treat any corruption as EOF
*/
if (binlog_can_be_corrupted &&
error != LOG_READ_MEM && error != LOG_READ_EOF)
{
my_b_seek(&log, prev_pos);
error=LOG_READ_EOF;
}
/*
TODO: now that we are logging the offset, check to make sure
the recorded offset and the actual match.
@ -1021,6 +1011,16 @@ end:
err:
thd_proc_info(thd, "Waiting to finalize termination");
if (my_errno == ER_MASTER_FATAL_ERROR_READING_BINLOG && my_b_inited(&log))
/*
detailing the fatal error message with coordinates
of the last position read.
*/
my_snprintf(error_text, sizeof(error_text), fmt, errmsg,
coord->file_name, (llstr(coord->pos, llbuff1), llbuff1),
log_file_name, (llstr(my_b_tell(&log), llbuff2), llbuff2));
else
strcpy(error_text, errmsg);
end_io_cache(&log);
RUN_HOOK(binlog_transmit, transmit_stop, (thd, flags));
/*
@ -1037,7 +1037,7 @@ err:
mysql_file_close(file, MYF(MY_WME));
thd->variables.max_allowed_packet= old_max_allowed_packet;
my_message(my_errno, errmsg, MYF(0));
my_message(my_errno, error_text, MYF(0));
DBUG_VOID_RETURN;
}