This is 4.4.0 but that milestone is missing.
When attempting to browse a certain table:
2015-04-02 13:31:18: (mod_fastcgi.c.2695) FastCGI-stderr: PHP Fatal error: Maximum execution time of 30 seconds exceeded in /srv/router/phpMyAdmin-4.4.0-english/libraries/dbi/DBIMysqli.class.php on line 594
2015-04-02 13:32:03: (mod_fastcgi.c.2695) FastCGI-stderr: PHP Fatal error: Maximum execution time of 30 seconds exceeded in /srv/router/phpMyAdmin-4.4.0-english/libraries/Util.class.php on line 295
2015-04-02 13:35:17: (mod_fastcgi.c.2695) FastCGI-stderr: PHP Fatal error: Maximum execution time of 30 seconds exceeded in /srv/router/phpMyAdmin-4.4.0-english/libraries/Util.class.php on line 302
The milestone was not missing but milestones are not presented in ascending order.
Oops. Is that a SF issue?
Yes, we don't control that.
Anything particular for this "certain table" ? Also, is it a case of using the "Show all" checkbox?
I'm using the nav tree, so no Show All checkbox. Table is nothing special, a MyISAM, 23 mb in 15700 rows.
Does it contain confidential data? If not, it would help to have access to a compressed export of this table so we can try to reproduce this.
I'll try to find out how to reproduce this.
It's executing this in a loop:
SELECT CONCAT(db_name, '.', table_name, '.', column_name) AS column_name, mimetype, transformation, transformation_options, input_transformation, input_transformation_options FROM phpmyadmin.pma__column_info WHERE db_name = 'test' AND table_name = 'users' AND ( mimetype != '' OR transformation != '' OR transformation_options != '' OR input_transformation != '' OR input_transformation_options != '');
How many columns in this table?
69 fields
This part of the code does not seem optimized. When I browse a table, asking for the default 25 rows, I see 25 calls to the PMA_getMIME() function which contains the SELECT CONCAT... you wrote about.
Does this correspond to what you are seeing?
2015-04-17 19:45 GMT+02:00 Marc Delisle lem9@users.sf.net:
Sounds like it.
Please try the attached patch.
Thanks, will try on Monday.
Tried it on another computer. It doesn't time out but it's still quite slow (like 5s).
Do you do any performance testing / tuning of pMA?
pMA 3 seems much faster than pMA 4 for me.
Thanks for testing the patch.
Browsing your table is 5s? Yes we try to tune performance but it's not easy replicating all kinds of user environments.
5s is still quite slow..
Unrelated to this bug ticket.
Not fixed yet..
20:51:07 write(6, "V\2\0\0\3SELECT CONCAT(
db_name
, '.',table_name
, '.', column_name`) AS column_name20:51:07 stat64("/srv/router/phpMyAdmin-4.4.3-english/libraries/special_schema_links.lib.php", {st_mode=S_IFREG|0644, st_size=16428, ...}) = 0
20:51:10 brk(0x9b76000) = 0x9b76000
<lots more="" brk()="">
20:51:23 brk(0x9cb6000) = 0x9cb6000
20:51:28 mmap2(NULL, 790528, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb12b6000
20:51:33 brk(0x9cf6000) = 0x9cf6000
20:51:37 --- SIGPROF {si_signo=SIGPROF, si_code=SI_KERNEL, si_value={int=1851877730, ptr=0x6e616962}} ---
This system is a bit older and slower, it's a AMD Athlon(tm) XP 2500+ with 512mb memory. Not sure what pMA is doing though.
Last edit: Marc Delisle 2015-04-20
30s, as it says in the error. ;)
Yes, still when attempting to browse.