REPOST: innobackex not backing up all data

Everyone - sorry for the repost.

I have a simple 15 GB DB and I when I run innobackupex it does not backup all the data in the ‘datadir’/mysql and 'datadir"/performance_schema directories.

After multiple runs it seems to not be consistent - sometimes it backups some of the data in those directories and sometimes it does not.

The output from the innobackupex says it did indeed back the data up - but this is clearly not the case. I am on ubuntu 11.10 VM running on KVM.

I have included two files: my.cnf and output from innobackex output where you can see that it says it backup the data in those files.

I am at wits end - any help would be appreciated.

Below is the “ls -l” in the db_backup directory showing the directories are empty:

total 16319044
-rw-r–r-- 1 root root 379 2012-07-30 15:02 backup-my.cnf
-rw-r----- 1 root root 5368709120 2012-07-30 15:09 ibdata1
-rw-r----- 1 root root 5368709120 2012-07-30 15:09 ibdata2
-rw-r----- 1 root root 5368709120 2012-07-30 15:09 ibdata3
-rw-r–r-- 1 root root 268435456 2012-07-30 15:09 ib_logfile0
-rw-r–r-- 1 root root 268435456 2012-07-30 15:09 ib_logfile1
drwxr-xr-x 2 root root 8192 2012-07-30 15:09 mysql
drwxr-xr-x 2 root root 4096 2012-07-30 15:09 performance_schema
-rw-r–r-- 1 root root 13 2012-07-30 15:09 xtrabackup_binary
-rw-r–r-- 1 root root 23 2012-07-30 15:09 xtrabackup_binlog_info
-rw-r----- 1 root root 77 2012-07-30 15:09 xtrabackup_checkpoints
-rw-r----- 1 root root 2097152 2012-07-30 15:09 xtrabackup_logfile

./mysql:
total 0

./performance_schema:
total 0

Hi,

As per the documentation: http://www.percona.com/doc/percona-xtrabackup/xtrabackup_bin /configuring.html?id=percona-xtrabackup:xtrabackup:configura tion

"We have noticed that NFS volumes not mounted with the sync option might not really sync the data. As a result, if you back up to an NFS volume mounted with the async option, and then try to prepare the backup from a different server that also mounts that volume, the data might appear to be corrupt. "

So I guess, this might be issue for you too. Can you use the sync mount option to avoid this problem and try again to take backup?

I do have the sync option on the NFS mount point options. So - that is not the issue. (IP addresses XX out).

Below are the NFS mount points for my server, you notice that the NFSVOL5 where the backups are has the sync option.

x.x.x.x:/vol/vFiler2amysqlVol1 /nfsvol1 nfs hard,rw,rsize=32768,wsize=32768,bg,vers=3,tcp,actimeo=600,no intr,suid,timeo=600,nolock 0 2
x.x.x.x:/vol/vFiler2bmysqlVol2 /nfsvol2 nfs hard,rw,rsize=32768,wsize=32768,bg,vers=3,tcp,actimeo=600,no intr,suid,timeo=600,nolock 0 2
x.x.x.x:/vol/vFiler2amysqlVol3 /nfsvol3 nfs hard,rw,rsize=32768,wsize=32768,bg,vers=3,tcp,actimeo=600,no intr,suid,timeo=600,nolock 0 2
x.x.x.x:/vol/vFiler2bmysqlVol4 /nfsvol4 nfs hard,rw,rsize=32768,wsize=32768,bg,vers=3,tcp,actimeo=600,no intr,suid,timeo=600,nolock 0 2
x.x.x.x:/vol/vFiler2amysqlVol5 /nfsvol5 nfs hard,rw,sync,rsize=32768,wsize=32768,bg,vers=3,tcp,actimeo=6 00,nointr,suid,timeo=600,nolock 0 2