Comments on: What is max_tmp_tables? http://ronaldbradford.com/blog/what-is-max_tmp_tables-2009-06-16/ Expert times and information on MySQL Fri, 12 Mar 2010 22:45:29 -0400 http://wordpress.org/?v=2.8.4 hourly 1 By: Sheeri http://ronaldbradford.com/blog/what-is-max_tmp_tables-2009-06-16/comment-page-1/#comment-573 Sheeri Wed, 17 Jun 2009 17:09:34 +0000 http://ronaldbradford.com/blog/?p=1645#comment-573 Paul is spot on (although see my latest post at <a HREF="http://www.pythian.com/news/2963/trivial-mysql-variable-of-the-day" rel="nofollow">http://www.pythian.com/news/2963/trivial-mysql-variable-of-the-day</a>, old_alter_table is not actually documented, and it's unclear whether or not changing the value does anything). I'll add that the word "yet" speaks to whether or not it did something before -- it did not -- currently it doesn't do anything "yet", which to me means it never has done anything. I've been combing the status variables recently (hence my posts; I did one last week on the silliness that is the "old" and "new" variables) and "this variable is unused" shows up in the docs in some places. So to find the variables that don't do anything, I used the search box on dev.mysql.com to search for "is unused" (and "do anything") in the MySQL Reference v5.1 docs only, clicked on "repeat the seach with the omitted results included", and found that I got many of the "options manual pages" and variables.... From http://dev.mysql.com/doc/refman/5.1/en/server-options.html --log-tc --tc-heuristic-recover From http://dev.mysql.com/doc/refman/5.1/en/server-status-variables.html Tc_log_max_pages_used Tc_log_page_size From http://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html the REFERENCES privilege From http://dev.mysql.com/doc/refman/5.1/en/replication-options-slave.html rpl_recovery_rank From http://dev.mysql.com/doc/refman/5.1/en/server-system-variables.html date_format datetime_format table_lock_wait_timeout time_format And a bunch in the API: http://dev.mysql.com/doc/refman/5.1/en/mysql-options.html http://dev.mysql.com/doc/refman/5.1/en/replication-options-slave.html#sysvar_slave_exec_mode I'll note for Paul -- max_tmp_tables is the only variable/option that uses the phrasing "does not yet do anything" instead of "is unused". Perhaps that should be fixed, for easier searching? Paul is spot on (although see my latest post at http://www.pythian.com/news/2963/trivial-mysql-variable-of-the-day, old_alter_table is not actually documented, and it’s unclear whether or not changing the value does anything). I’ll add that the word “yet” speaks to whether or not it did something before — it did not — currently it doesn’t do anything “yet”, which to me means it never has done anything.

I’ve been combing the status variables recently (hence my posts; I did one last week on the silliness that is the “old” and “new” variables) and “this variable is unused” shows up in the docs in some places. So to find the variables that don’t do anything, I used the search box on dev.mysql.com to search for “is unused” (and “do anything”) in the MySQL Reference v5.1 docs only, clicked on “repeat the seach with the omitted results included”, and found that I got many of the “options manual pages” and variables….

From http://dev.mysql.com/doc/refman/5.1/en/server-options.html
–log-tc
–tc-heuristic-recover

From http://dev.mysql.com/doc/refman/5.1/en/server-status-variables.html
Tc_log_max_pages_used
Tc_log_page_size

From http://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html
the REFERENCES privilege

From http://dev.mysql.com/doc/refman/5.1/en/replication-options-slave.html
rpl_recovery_rank

From http://dev.mysql.com/doc/refman/5.1/en/server-system-variables.html
date_format
datetime_format
table_lock_wait_timeout
time_format

And a bunch in the API:
http://dev.mysql.com/doc/refman/5.1/en/mysql-options.html
http://dev.mysql.com/doc/refman/5.1/en/replication-options-slave.html#sysvar_slave_exec_mode

I’ll note for Paul — max_tmp_tables is the only variable/option that uses the phrasing “does not yet do anything” instead of “is unused”. Perhaps that should be fixed, for easier searching?

]]> By: Paul DuBois http://ronaldbradford.com/blog/what-is-max_tmp_tables-2009-06-16/comment-page-1/#comment-574 Paul DuBois Wed, 17 Jun 2009 11:21:48 +0000 http://ronaldbradford.com/blog/?p=1645#comment-574 "If this option doesn’t yet do anything, why is it there?" Because it's visible in the server (e.g., with SHOW VARIABLES). If it weren't documented, people would ask "why is it not listed in the manual?" "Does it actually do something and the documentation is misrepresenting the option?" No. "Did it provide some feature or functionality before?" Not that I know of. Perhaps it was originally added in anticipation of adding a feature that ended up not being developed. That's only speculation on my part, though. "The latest is max_tmp_tables." It was added in 3.23.0, and it's been in the manual as described for quite a long time. “If this option doesn’t yet do anything, why is it there?”

Because it’s visible in the server (e.g., with SHOW VARIABLES). If it weren’t documented, people would ask “why is it not listed in the manual?”

“Does it actually do something and the documentation is misrepresenting the option?”

No.

“Did it provide some feature or functionality before?”

Not that I know of. Perhaps it was originally added in anticipation of adding a feature that ended up not being developed. That’s only speculation on my part, though.

“The latest is max_tmp_tables.”

It was added in 3.23.0, and it’s been in the manual as described for quite a long time.

]]>