Peter Skipworth writes:
> Hi all,
>
> Has anyone experienced anything like the following ?
>
> I have several queries which, admittedly, are quite complex and operate on
> several million rows of indexed data. The tables are naturally locked
> while the query occurs, which is fine, but what is *not* fine is that
> occasionally, completely unrelated tables end up needing a repair after
> the long query has completed.
>
> The only pattern I can see is that towards the end of the query, I see a
> hell of a lot of "Opening table" & "Closing table" status messages in the
> processlist - as soon as I see this, I can assume that the mentioned
> tables will have become corrupted - see a sample processlist below.
> Any ideas ?
>
> Cheers,
>
> P
>
>
>
> 164804 | hot | localhost | hot | Query | 6 | closing tables |
> select * from tbl_hotclients limit 0
> |
> | 164805 | hot | localhost | hot | Query | 6 | closing tables |
> select * from tbl_hotclients limit 0
> |
> | 164808 | hot | localhost | hot | Query | 14 | closing tables |
> select job_id from tbl_job_events where session='98760270421222662058432'
> and event_type='view' and |
> | 164809 | hot | localhost | hot | Query | 14 | closing tables |
> select tbl_jobs.refresh_date, tbl_jobs.job_id, tbl_jobs.title,
> tbl_jobs.region_id, tbl_jobs.hl_regio |
> | 164811 | hot | localhost | hot | Query | 5 | Opening tables |
> select * from tbl_homepage_crazyjobs limit 0
> |
> | 164823 | hot | localhost | hot | Query | 12 | closing tables |
> select * from tbl_hours where job_hours_id = 7
> |
> | 164824 | hot | localhost | hot | Query | 6 | closing tables |
> select * from tbl_banner_server limit 0
> |
> | 164825 | hot | localhost | hot | Query | 54 | Locked |
> insert into tbl_webevents_temp
> (processtime,event_time,cookie,parent,referer,duration,newuser,a,job_ |
> | 164832 | hot | localhost | hot | Query | 10 | closing tables |
> select * from tbl_applications limit 0
> |
> | 164833 | hot | localhost | hot | Query | 5 | Opening tables |
> select * from tbl_applications limit 0
> |
> | 164834 | hot | localhost | hot | Query | 6 | closing tables |
> select * from tbl_applications limit 0
> |
> | 164841 | hot | localhost | hot | Query | 10 | closing tables |
> select * from tbl_application_emails limit 0
> |
> | 164842 | hot | localhost | hot | Query | 5 | Opening table |
> insert into tbl_job_events (job_id, session, event_type, user_type,
> recorded, event_time, uni_id) va |
>
>
> -----------------------------------------------
> Peter Skipworth Perl Developer/Unix Systems Administrator
> [EMAIL PROTECTED]
> +44 795 055 0029
> -----------------------------------------------
> 4c 61 20 6c 6f 79 52 75 62 20 73 41 20 65 52 61
> 20 45 65 62 4f 4c 47 6e 74 20 20 4f 53 75 00 0a
>
Hi!
What we need is a repeatable test case, i.e. a set of tables and
script that will always lead to table corruption.
If you can come up with such a test case, forward it to
[EMAIL PROTECTED]
Regards,
Sinisa
____ __ _____ _____ ___ == MySQL AB
/*/\*\/\*\ /*/ \*\ /*/ \*\ |*| Sinisa Milivojevic
/*/ /*/ /*/ \*\_ |*| |*||*| mailto:[EMAIL PROTECTED]
/*/ /*/ /*/\*\/*/ \*\|*| |*||*| Larnaca, Cyprus
/*/ /*/ /*/\*\_/*/ \*\_/*/ |*|____
^^^^^^^^^^^^/*/^^^^^^^^^^^\*\^^^^^^^^^^^
/*/ \*\ Developers Team
---------------------------------------------------------------------
Before posting, please check:
http://www.mysql.com/manual.php (the manual)
http://lists.mysql.com/ (the list archive)
To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php