Merge lp:~gl-az/percona-xtrabackup/2.1-bug1197644 into lp:percona-xtrabackup/2.1

[George Ormond Lorch III]

Fix for bug 1197644 : --encrypt-key-file is not covered by the test suite.

Modified xb_encrypt.sh to test using the --encrypt-key-file option while other tests that use encryption will use --encrypt-key option.

[George Ormond Lorch III (gl-az)]

Jenkins run on only xb_encrypt.sh

http://jenkins.percona.com/view/XtraBackup/job/percona-xtrabackup-2.1-param/369/

[Alexey Kopytov (akopytov)]

Using global files (under /tmp) in tests is a bad idea, because 1) there may be a conflict with other concurrently running tests and 2) they won't be removed if the test fails.

If you create them under $TEST_VAR_ROOT they will be unique for each test and will be automatically removed by the test suite if the test does not clean up itself.

review: Needs Fixing

[George Ormond Lorch III (gl-az)]

Yeah, I thought about that yesterday after I submitted the MP and
remembered that my original test for this was doing something like that
for the exact same reason. I was hoping that you wouldn't get to the
review until I had a chance to kill that MP and rework it. The problem
is that this is using the xb_local functionality so it means that
instead of just including xb_local I will have to actually change this
to be a full test so that $TEST_VAR_ROOT is properly defined by the time
the script goes to use it. I'll fix this up today.

On 7/8/2013 11:10 PM, Alexey Kopytov wrote:
> Review: Needs Fixing
>
> Using global files (under /tmp) in tests is a bad idea, because 1) there may be a conflict with other concurrently running tests and 2) they won't be removed if the test fails.
>
> If you create them under $TEST_VAR_ROOT they will be unique for each test and will be automatically removed by the test suite if the test does not clean up itself.

--
George O. Lorch III
Software Engineer, Percona
[phone] x542 US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

[Alexey Kopytov (akopytov)]

George,

TEST_VAR_ROOT is defined for each test, you don't have to include
anything to have it defined.

[George Ormond Lorch III (gl-az)]

Ahh, awesome. I was just looking at some of the exported vars when I
hung it up yesterday while doing the code review for the test suite
cleanup and didn't connect the dots. That will make it quite easy then...
On 7/9/2013 7:28 AM, Alexey Kopytov wrote:
> George,
>
> TEST_VAR_ROOT is defined for each test, you don't have to include
> anything to have it defined.
>

--
George O. Lorch III
Software Engineer, Percona
[phone] x542 US/Arizona (GMT -7)
skype: george.ormond.lorch.iii

[George Ormond Lorch III (gl-az)]

http://jenkins.percona.com/view/XtraBackup/job/percona-xtrabackup-2.1-param/372/

[Alexey Kopytov (akopytov)]

review: Approve

=== modified file 'test/t/xb_encrypt.sh'
--- test/t/xb_encrypt.sh	2013-07-09 15:32:43 +0000
+++ test/t/xb_encrypt.sh	2013-07-09 15:32:43 +0000
@@ -4,13 +4,18 @@
 
 encrypt_algo="AES256"
 encrypt_key="percona_xtrabackup_is_awesome___"
-
-innobackupex_options="--encrypt=$encrypt_algo --encrypt-key=$encrypt_key --encrypt-threads=4 --encrypt-chunk-size=8K"
+encrypt_key_file=${TEST_VAR_ROOT}/xb_encrypt.key
+
+echo -n $encrypt_key > $encrypt_key_file
+
+innobackupex_options="--encrypt=$encrypt_algo --encrypt-key-file=$encrypt_key_file --encrypt-threads=4 --encrypt-chunk-size=8K"
 data_decrypt_cmd="for i in *.xbcrypt; do \
-xbcrypt -d -a $encrypt_algo -k $encrypt_key < \$i > \${i:0:\${#i}-8}; \
+xbcrypt -d -a $encrypt_algo -f $encrypt_key_file < \$i > \${i:0:\${#i}-8}; \
 rm -f \$i; done; \
 for i in ./sakila/*.xbcrypt; do \
-xbcrypt -d -a $encrypt_algo -k $encrypt_key < \$i > \${i:0:\${#i}-8}; \
+xbcrypt -d -a $encrypt_algo -f $encrypt_key_file < \$i > \${i:0:\${#i}-8}; \
 rm -f \$i; done;"
 
 . inc/xb_local.sh
+
+rm -f $encrypt_key_file