Details
Description
Build command:
cmake . -DCMAKE_BUILD_TYPE=Debug -DWITH_VALGRIND=1
make -j `cat /proc/cpuinfo | grep processor| wc -l`
./mtr encryption.innodb-discard-import got fails:
worker[1] Using MTR_BUILD_THREAD 300, with reserved ports 16000..16019
encryption.innodb-discard-import 'cbc,innodb' [ fail ]
Test ended at 2017-01-25 01:41:23
CURRENT_TEST: encryption.innodb-discard-import
— /u01/MariaDB/mysql-test/suite/encryption/r/innodb-discard-import.result 2017-01-25 00:54:30.909246787 +0800
+++ /u01/MariaDB/mysql-test/suite/encryption/r/innodb-discard-import.reject 2017-01-25 01:41:23.361297842 +0800
@@ -129,6 +129,6 @@
- t3 yes on expecting NOT FOUND
NOT FOUND /tmpres/ in t3.ibd - t4 yes on expecting NOT FOUND
-NOT FOUND /mysql/ in t4.ibd
+FOUND /mysql/ in t4.ibd
DROP PROCEDURE innodb_insert_proc;
DROP TABLE t1,t2,t3,t4;
Attachments
Issue Links
- relates to
-
MDEV-16116 encryption.innodb_encryption_discard_import still fails in buildbot with FOUND vs NOT FOUND
-
- Closed
-
Activity
I just got this on 10.4:
CURRENT_TEST: encryption.innodb_encryption_discard_import
— /home/my/maria-10.4/mysql-test/suite/encryption/r/innodb_encryption_discard_import.result 2022-10-04 10:51:31.191518134 +0300
+++ /home/my/maria-10.4/mysql-test/suite/encryption/r/innodb_encryption_discard_import.reject 2023-05-04 16:10:34.087764702 +0300
@@ -22,7 +22,7 @@
- t1 yes on expecting NOT FOUND
NOT FOUND /foobar/ in t1.ibd - t2 ... on expecting NOT FOUND
-NOT FOUND /temp/ in t2.ibd
+FOUND 1 /temp/ in t2.ibd - t3 ... on expecting NOT FOUND
NOT FOUND /barfoo/ in t3.ibd - restart
This test fails in the same way across many versions (10.4 onwards). On 10.6+ there are some crashes on POWER due to a Linux kernel bug (MDEV-30728).
monty, the test encryption.innodb_encryption_discard_import is different from encryption.innodb-discard-import. The encryption.innodb-discard-import test seems to be stable also according to the cross-reference (if we ignore MDEV-30728). There were some hangs of the test, the last one on 2022-09-12 on s390x and 2021-10-20 on AMD64.
My attempts to reproduce a failure of encryption.innodb_encryption_discard_import have failed so far, but I will keep trying.
The test encryption.innodb_encrypt_discard_import is creating unnecessarily large tables and restarting the server for no good reason. I simplified the test to reduce its execution time from more than 20 seconds to less than 2 seconds on my system. Hopefully this make the failures disappear.
MDEV-16116 had been filed for the failure of the test encryption.innodb_encryption_discard_import.
commit 17430a802b4867c60e7185807c0e2ae14c878475
Author: Jan Lindström <jan.lindstrom@mariadb.com>
Date: Wed Jan 25 08:43:19 2017 +0200
MDEV-11905: encryption.innodb-discard-import test failsProblem was that on import with buffered input/output
trim is not possible as we write several pages in
one write.