I also have this error when I run a remote backup with Innobackupex, and the result is that the xtrabackup_log is missing:

One the box the backup is run from:

110713 13:54:13 innobackupex: Finished backing up .frm, .MRG, .MYD, .MYI, .TRG, .TRN, .ARM, .ARZ, .CSV, .CSM and .opt files

innobackupex: Resuming ibbackup

xtrabackup: The latest check point (for incremental): '3497781933050'
>> log scanned up to (3498117720317)
xtrabackup: Transaction log of lsn (3492260936252) to (3498117720317) was copied.
innobackupex: Error: mysql child process has died: ERROR 2006 (HY000) at line 15: MySQL server has gone away
while waiting for reply to MySQL request: 'UNLOCK TABLES;' at /usr/bin/innobackupex line 336.

On the box that backup is made to:

total 215050336
-rw-r--r-- 1 root root 353 2011-07-13 13:50 backup-my.cnf
drwxr-xr-x 2 root root 4096 2011-07-13 19:04 gearman
-rw-r----- 1 root root 220211445760 2011-07-13 13:48 ibdata1
drwxr-xr-x 2 root root 20480 2011-07-13 19:05 mail
drwxr-xr-x 2 root root 20480 2011-07-13 19:05 mail_test
drwxr-xr-x 2 root root 4096 2011-07-13 19:05 mysql
drwxr-xr-x 2 root root 4096 2011-07-13 19:05 weblogs
-rw-r--r-- 1 root root 28 2011-07-13 19:04 xtrabackup_binlog_info
-rw-r--r-- 1 root root 77 2011-07-13 19:04 xtrabackup_slave_info

No xtrabackup_logfile is present due to InnoBackupex error above causing the backup to fail.

Line 695 of innobackupex:

sub wait_for_ibbackup_suspend {
print STDERR "$prefix Waiting for ibbackup (pid=$ibbackup_pid) to suspend\n";
print STDERR "$prefix Suspend file '$suspend_file'\n\n";
for (; {
sleep 2;
last if -e $suspend_file;

# check that ibbackup child process is still alive
if ($ibbackup_pid == waitpid($ibbackup_pid, &WNOHANG)) {
$ibbackup_pid = '';
Die "ibbackup child process has died";
}
}

So, xtrabackup child process somehow died. ?

try changing the wait_timeout value on the server to a much larger value. This could be caused by server killing the backup connection since wait_timeout was exceeded.

looks like the wait_timeout parameter is not set to a value greater than the backup time, hence mysql is killing the connection.

P.S sorry if this is a double post, i already submitted a reply and that seems to have disappeared Replication issues ?

Thank you! I will try this.

Initially i had 120 as wait_timeout. Increasing it to 1200 didnt make any change. Will try a really higher value next night in a maintanance window.