Testing XB branch ~laurynas-biveinis/percona-xtrabackup/BT-16274-xtrabackup-changed-page-tracking against PS branch ~laurynas-biveinis/percona-server/QA-16724-5.5-2
Use my.cnf:
[client]
port=5528
socket=/data/dev/16274/2/1/mysql.sock
[mysqld]
basedir=/usr/local/dev/ps-5.5-16274
datadir=/data/dev/16274/2/1/data
tmpdir=/data/dev/16274/2/1/tmp
port=5528
socket=/data/dev/16274/2/1/mysql.sock
pid-file=/data/dev/16274/2/1/mysql.pid
console
user=root
server-id=1
innodb_track_changed_pages=TRUE
innodb_buffer_pool_size=2G
innodb_log_file_size=500M
Start server and RQG against server with percona_qa grammer.
Perform a baseline 'full' backup:
innobackupex --defaults-file=... --ibbackup=xtrabackup --no-timestamp ./0
then loop, creating successive incrementals:
for (( i=1; i < 20; i++ ))
do
sleep 40 # or 10 or 20 or 30, depends on machine speed and debug, with full debug I need 40, in release 10 is enough
local prev=`expr $i - 1`
innobackupex --defaults-file=... --ibbackup=xtrabackup --no-timestamp --incremental ./$i --incremental-basedir=./$prev
done
Shutdown the server.
Now start the restoration process:
mkdir ./working
innobackupex --defaults-file=... --ibbackup=xtrabackup --apply-log --redo-only ./0
innobackupex --defaults-file=... --ibbackup=xtrabackup --apply-log --redo-only ./0 --incremental-dir=./1
innobackupex --defaults-file=... --ibbackup=xtrabackup --apply-log --redo-only ./0 --incremental-dir=./2
etc...
Along this process, you will get an assertion from xtrabackup:
InnoDB: Progress in percents: 0 1 2 3 4 5 6 7 130201 19:03:11 InnoDB: Assertion failure in thread 140550848112384 in file rem0rec.c line 561
InnoDB: We intentionally generate a memory trap.
InnoDB: Submit a detailed bug report to http://bugs.mysql.com.
InnoDB: If you get repeated assertion failures or crashes, even
InnoDB: immediately after the mysqld startup, there may be
InnoDB: corruption in the InnoDB tablespace. Please refer to
InnoDB: http://dev.mysql.com/doc/refman/5.5/en/forcing-innodb-recovery.html
InnoDB: about forcing recovery.
innobackupex: Error:
innobackupex: ibbackup failed at /data/bin/xb-2.1-16274-5.5/innobackupex line 383.
core backtrace:
#0 0x000000391a8328a5 in raise () from /lib64/libc.so.6
#1 0x000000391a834085 in abort () from /lib64/libc.so.6
#2 0x00000000004c2f71 in rec_get_offsets_func (rec=0x7fd4819f00bb "", index=0x7fd478001e28, offsets=0x7fd48b5fd480, n_fields=18446744073709551615, heap=0x7fd48b5fd7a8,
file=0x592168 "/home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/page/page0cur.c", line=1850) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/rem/rem0rec.c:561
#3 0x00000000004ae87d in page_cur_parse_delete_rec (ptr=0x7fd484b5460a "", end_ptr=0x7fd484b5460a "", block=0x7fd48099dc80, index=0x7fd478001e28, mtr=0x7fd48b5fd880) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/page/page0cur.c:1849
#4 0x0000000000499e67 in recv_parse_or_apply_log_rec_body (type=42 '*', ptr=0x7fd484b54608 "", end_ptr=0x7fd484b5460a "", block=0x7fd48099dc80, mtr=0x7fd48b5fd880) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/log/log0recv.c:1247
#5 0x000000000049a959 in recv_recover_page_func (just_read_in=1, block=0x7fd48099dc80) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/log/log0recv.c:1666
#6 0x00000000004402cb in buf_page_io_complete (bpage=0x7fd48099dc80) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/buf/buf0buf.c:4065
#7 0x000000000046f19f in fil_aio_wait (segment=5) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/fil/fil0fil.c:5739
#8 0x00000000004f7cd7 in io_handler_thread (arg=0xa78348) at /home/glorch/dev/xb-2.1-16274-5.5/Percona-Server-5.5/storage/innobase/srv/srv0start.c:485
#9 0x000000391b007851 in start_thread () from /lib64/libpthread.so.0
#10 0x000000391a8e811d in clone () from /lib64/libc.so.6
core dump