innobackupex fails to backup CSV tables causing mysql_upgrade to fail
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Percona XtraBackup moved to https://jira.percona.com/projects/PXB |
Fix Released
|
High
|
Valentine Gostev |
Bug Description
In MySQL 5.1, there are two CSV tables in the mysql schema (general_log and slow_log) which are not backuped by innobackupex.pl at all. So when the backup is used to build a slave and 'mysql_upgrade' is run, you get the following error:
<pre>
mysql.time_zone OK
mysql.time_
mysql.time_
mysql.time_
mysql.time_
mysql.user OK
Running 'mysql_
ERROR 13 (HY000) at line 311: Can't get stat of './mysql/
ERROR 13 (HY000) at line 316: Can't get stat of './mysql/
FATAL ERROR: Upgrade failed
</pre>
Solution:
Patch innobackupex.pl to include CSV tables, specifically line 1811 from
my $wildcard = '*.{frm,
to
my $wildcard = '*.{frm,
Related branches
- Vadim Tkachenko: Approve
- Alexey Kopytov (community): Approve
-
Diff: 700 lines (+208/-107)11 files modifiedinnobackupex (+7/-7)
test/inc/common.sh (+15/-0)
test/t/ib_csm_csv.sh (+63/-0)
test/t/ib_include.sh (+46/-0)
test/t/ib_incremental.sh (+12/-16)
test/t/ib_specialchar.sh (+13/-13)
test/t/xb_basic.sh (+12/-12)
test/t/xb_incremental.sh (+9/-15)
test/t/xb_parallel.sh (+14/-14)
test/t/xb_partial.sh (+5/-19)
test/t/xb_stream.sh (+12/-11)
Changed in percona-xtrabackup: | |
milestone: | none → release-1.2 |
assignee: | nobody → Aleksandr Kuzminsky (akuzminsky) |
Changed in percona-xtrabackup: | |
milestone: | 1.2 → 1.3.1 |
Changed in percona-xtrabackup: | |
assignee: | Aleksandr Kuzminsky (akuzminsky) → Alexey Kopytov (akopytov) |
importance: | Undecided → Medium |
milestone: | 1.3.1 → 1.6 |
status: | New → Opinion |
Changed in percona-xtrabackup: | |
importance: | Medium → High |
Changed in percona-xtrabackup: | |
assignee: | Alexey Kopytov (akopytov) → Valentine Gostev (core-longbow) |
Changed in percona-xtrabackup: | |
status: | Opinion → In Progress |
Changed in percona-xtrabackup: | |
status: | In Progress → Fix Released |
status: | Fix Released → Fix Committed |
Changed in percona-xtrabackup: | |
status: | Fix Committed → Fix Released |
Why is this bug being downgraded and put off to later milestones? I'd say that for a backup program to fail to backup something it is supposed to is about as serious as it could possibly get? Especially since restoring from a backup results in a partially corrupt and apparently unfixable database. Also it seems that the posted fix is completely trivial - are there other factors preventing it from being used?