| Bug #38811 | Test main.variables-big failing (all versions) | ||
|---|---|---|---|
| Submitted: | 14 Aug 2008 20:52 | Modified: | 4 Sep 2008 17:41 |
| Reporter: | Patrick Crews | Email Updates: | |
| Status: | Closed | Impact on me: | |
| Category: | Tests: Server | Severity: | S3 (Non-critical) |
| Version: | 5.0+ | OS: | Any |
| Assigned to: | Patrick Crews | CPU Architecture: | Any |
| Tags: | big test, test, test failure | ||
[14 Aug 2008 22:16]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/51667 2662 Patrick Crews 2008-08-14 Bug#38811: Test main.variables-big failing (all versions) Modified test to remove SHOW PROCESSLIST (too variable, overkill to check for server crash) Re-recorded .result file to account for truncate warnings and new SQL statements.
[4 Sep 2008 17:41]
Patrick Crews
This bug was fixed as part of the patch for Bug#32124 crash if prepared statements refer to variables in the where clause Closing out this bug as the test was patched and works as expected now.
Description: Test main.variables-big is failing on all versions (for various reasons) On 5.0 - failing because show processlist's actual ID column values are different from the expected. On 5.1 - failing because expected results do not contain the warnings that pop up when transaction_prealloc_size is truncated to the max. limit (4294966272) On 6.0 - same as 5.1 How to repeat: From the mysql-test directory, run the following: ./mysql-test-run.pl --force --big --do-test variables-big You will observe different mismatches depending on the server version Suggested fix: Update the test to run accordingly: 1) We can limit the number of runs to the following: a. One run < max limit (4294966272) b. One run > max limit to verify truncation c. Possibly one run setting it to 0 2) We can eliminate show processlist and replace this with any SQL statement to verify that the server hasn't crashed. (This test was created for Bug #27322 failure to allocate transaction_prealloc_size causes crash) Probably, something to examine the value of transaction_prealloc_size.