RE: mysql-server-5.1.19 path variable error set on compile

2007-06-18 Thread Geoffroy Cogniaux
Hi,

Set this variable at command line or in your my.cnf.

Geoffroy.

-Message d'origine-
De : David Southwell [mailto:[EMAIL PROTECTED] 
Envoyé : dimanche 17 juin 2007 17:33
À : mysql@lists.mysql.com
Objet : mysql-server-5.1.19 path variable error set on compile


# uname -a
 6.1-RELEASE FreeBSD 6.1-RELEASE #0: Sun May  7 04:15:57 UTC 2006 
[EMAIL PROTECTED]:/usr/obj/usr/src/sys/SMP  amd64
#
Just upgraded mysql to include:

# pkg_info |grep mysql
mysql-client-5.1.19 Multithreaded SQL database (client)
mysql-server-5.1.19 Multithreaded SQL database (server)
p5-DBD-mysql51-4.005 MySQL 5.1 driver for the Perl5 Database Interface (DBI)
php5-mysql-5.2.3The mysql shared extension for php
#

Seems we have an oddity here after compiling:
Extract from show variables:
show variables;
+-+---+
| Variable_name   | Value |
+-+---+
| basedir | /usr/local/  
| character_sets_dir  | /usr/local/share/mysql/charsets/  |
   
| datadir | /usr2/datadb/ |

| general_log_file| /usr2/datadb/dns1.log |

| language| /usr/local/share/mysql/english/   |

| pid_file| /usr2/datadb//dns1.vizion2000.net.pid |

 ^^Why '//'

| plugin_dir  | /usr/local/lib/mysql  |

| slow_query_log_file | /usr2/datadb/dns1-slow.log|
| socket  | /tmp/mysql.sock   |

So I thought I would:
(a) RESET the variable

mysql set pid_file = /usr2/datadb/dns1.vizion2000.net.pid ;
ERROR 1193 (HY000): Unknown system variable 'pid_file'
mysql set GLOBAL pid_file = /usr2/datadb/dns1.vizion2000.net.pid ;
ERROR 1193 (HY000): Unknown system variable 'pid_file'
mysql set GLOBAL pid_file = '/usr2/datadb/dns1.vizion2000.net.pid' ;
ERROR 1193 (HY000): Unknown system variable 'pid_file'
mysql set @@GLOBAL pid_file = '/usr2/datadb/dns1.vizion2000.net.pid' ;
ERROR 1064 (42000): You have an error in your SQL syntax; check the manual 
that corresponds to your MySQL server version for the right syntax to use 
near 'pid_file = '/usr2/datadb/dns1.vizion2000.net.pid'' at line 1
mysql set @@GLOBAL.pid_file = '/usr2/datadb/dns1.vizion2000.net.pid' ;
ERROR 1193 (HY000): Unknown system variable 'pid_file'
mysql  

But I cannot seem to get the Syntax right - can someone please point me in
the 
right direction.

(b) Ask why this might be happening.

It might be worth recording that /usr2/ is a seperate physical device and 
emphasize that the base-dir is /usr/local/.

Thanks in advance

David 


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Re: Can I undelete a dropped MyISAM table?

2007-06-17 Thread Geoffroy Cogniaux

2007/6/16, mos [EMAIL PROTECTED]:

At 09:39 AM 6/16/2007, Stanley wrote:
Is it possible to undelete a dropped MyISAM table?

Nope, not from within MySQL. Your operating system might allow you to
undelete the file then you might have a chance of importing the data to a
new table.

Mike



If your have a backup of your file system, you have only to restore
those file in the mysql datadir : yourtable.FRM, yourtable.MYD,
yourtable.MYI.
No need to importing data to a new table. MySQL will see it as it were
before dropping it.
This is a notable difference between myisam and innodb for instance.

Regards,
Geoffroy.

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Re: Installation problems with MySql 5.0.41 (source distribution)

2007-06-17 Thread Geoffroy Cogniaux

Hi,

You're running MySQL with a mysql linux user, verify that he have
write access on mysql datadir. You should specify it with --datadir=
at command line invoking mysql_install_db too.

2007/6/16, [EMAIL PROTECTED] [EMAIL PROTECTED]:


Hi everybody,

I am trying to compile/configure MySQl 5.0.41 on a Mandrake 10 linux box.

In doing so, I am getting some errors with mysql_install_db (ERROR: 1049
Unknown database 'mysql', Installation of system tables failed!)

Please let me know how to solve the problem.

Thanks in advance for your time and help.

Anand

Here are the details pertaining to my problem:

I used this as a guide:
http://dev.mysql.com/doc/refman/5.0/en/quick-install.html:

and followed all the steps. The following error(s) comes when I try using
mysql_install_db

/usr/local/mysql/bin/mysql_install_db --user=mysql
Installing MySQL system tables...
ERROR: 1049  Unknown database 'mysql'
070616  1:24:38 [ERROR] Aborting
070616  1:24:38 [Note] /usr/local/mysql/libexec/mysqld: Shutdown complete
Installation of system tables failed!

Examine the logs in /usr/local/mysql/var for more information.
You can try to start the mysqld daemon with:
/usr/local/mysql/libexec/mysqld --skip-grant 
and use the command line tool
/usr/local/mysql/bin/mysql to connect to the mysql
database and look at the grant tables:

shell /usr/local/mysql/bin/mysql -u root mysql
mysql show tables

Try 'mysqld --help' if you have problems with paths. Using --log
gives you a log in /usr/local/mysql/var that may be helpful.
---

And then, when I do /usr/local/mysql/libexec/mysqld --skip-grant  as
suggested in the above error message, it just aborts.

Here is the architecture info. from mysqlbug
--
Release:   mysql-5.0.41 (Source distribution)
C compiler:gcc (GCC) 3.3.2 (Mandrake Linux 10.0 3.3.2-6mdk)
C++ compiler:  g++ (GCC) 3.3.2 (Mandrake Linux 10.0 3.3.2-6mdk)
Environment:
machine, os, target, libraries (multiple lines)
System: Linux 2.6.3-7mdk #1 Wed Mar 17 15:56:42 CET 2004 i686 unknown
unknown GNU/Linux
Architecture: i686

Some paths:  /usr/bin/perl /usr/bin/make /usr/bin/gmake /usr/bin/gcc
/usr/bin/cc
GCC: Reading specs from /usr/lib/gcc-lib/i586-mandrake-linux-gnu/3.3.2/specs
Configured with: ../configure --prefix=/usr --libdir=/usr/lib
--with-slibdir=/lib --mandir=/usr/share/man --infodir=/usr/share/info
--enable-shared --enable-threads=posix --disable-checking
--enable-long-long --enable-__cxa_atexit --enable-clocale=gnu
--enable-languages=c,c++,ada,f77,objc,java,pascal
--host=i586-mandrake-linux-gnu --with-system-zlib
Thread model: posix
gcc version 3.3.2 (Mandrake Linux 10.0 3.3.2-6mdk)
Compilation info: CC='gcc'  CFLAGS=''  CXX='g++'  CXXFLAGS=''  LDFLAGS=''
ASFLAGS=''
LIBC:
lrwxrwxrwx  1 root root 13 Sep 13  2004 /lib/libc.so.6 - libc-2.3.3.so
-rwxr-xr-x  1 root root 1281788 Feb 16  2004 /lib/libc-2.3.3.so
-rw-r--r--  1 root root 204 Feb 16  2004 /usr/lib/libc.so
Configure command: ./configure '--prefix=/usr/local/mysql' '--with-unix\
-socket-path=/usr/local/mysql/tmp/mysql.sock'
--


--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Re: mysqldump for myisam tables.

2007-06-17 Thread Geoffroy Cogniaux

2007/6/16, ViSolve DB Team [EMAIL PROTECTED]:

-- this itself takes consistent backup. mysqldump utility by default locks
the table.

yes with --lock-tables option which do a FLUSH TABLES WITH READ LOCK
and is a good consistent way for myisam tables.

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: Sorting by a list of possible results in a column....

2007-06-05 Thread Geoffroy Cogniaux
Hi,

You could use a case statement to implicitly convert your column to
everything you want:

select status, 
   case when status = 'undefined' then 4
   when status = 'Top Priority' then 1
   ... 
   End as ord_status
from development
order by ord_status

Bye
Geoffroy

-Message d'origine-
De : Mike Morton [mailto:[EMAIL PROTECTED] 
Envoyé : mardi 5 juin 2007 23:26
À : mysql@lists.mysql.com
Objet : Sorting by a list of possible results in a column 

I am sure that this has been asked - but in searching through google and
lists for about an hour - hopefully someone will indulge me a repeat
question here.

I have a query that selects a list of results, ordering them by the status
field.  However, I want to further sort that by the type of status, that is:

Undefined
Ready for Review
Top Priority
Priority
Completed
Etc...

Every sort that I try, of course, sorts alphabetically.  Is there a way to
define how the sort function works in the order by?

I know that I could do this in PHP after populating the results into an
array, but that is (in my opinion) an unnecessary step that could be handled
at the database query

Thanks!

Sample query:
Select * from development order by status


-- 
Cheers

Mike Morton


*
* Tel: 905-465-1263
* Email: [EMAIL PROTECTED]
*





-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: MySQL 5.0 creates 100,000 Window handles

2007-06-02 Thread Geoffroy Cogniaux
Hi,

In fact, this is a good question. Therefore, I gonna try to answer.

 1) What is MySQL using the handles for?
Handles are nothing and all on Windows. Nothing because they are only
pointers to hidden internal struct. And all, because, Handle are everywhere
if you try to develop Win32 App.
There are 3 kinds:
Users: Window, Cursor, Menu,...
GDI: all graphic objects such as Brush, Pen,...
Kernel: Access token (ACL), Console input, Event, File, Heap, Mutex, Pipe,
Process, Semaphore, Socket, Thread, Timer, ... 
A database server is therefore a great consumer of kernel Handles.

 4) How does a MySQL server handle millions of queries on a large table?
 Will it hit an upper limit of Handles that it can allocate?
The per-process theoretical limit on kernel handles is 2^24. However,
handles are stored in the paged pool (kernel reserved memory), so the actual
number of handles you can create is based on available memory. So, the
number of handles that you can create on 32-bit Windows is significantly
lower than 2^24. Example on Windows2000, max pool size is 300 MB (I don't
know on most recent windows versions).
But be sure MySQL server can handle millions of queries on a large table.

 2) Is there any way to stop MySQL from consuming so many Windows
 resources?
I investigated and can summarize with a simple test I made with a 5.0.37
compiled with all storage engines (and verified with a 4.1.21): MySQL
allocates 43000 handles. I recompiled it without InnoDB and BDB, and MySQL
allocates now 108 handles at startup. I'm not an indeep Innodb's
specialist neither BDB, but I know they have row locking mechanism in
difference of MyIsam but I'm sure that they are great consumers of Mutex.

 Windows of course runs slow with this many handles allocated.
Yes, not because of number of handles (logical resources) but because of
physical resources, especially RAM, and perhaps by-design in OS kernel.


Regards,
Geoffroy

-Message d'origine-
De : Geoffroy Cogniaux [mailto:[EMAIL PROTECTED] 
Envoyé : dimanche 27 mai 2007 23:13
À : 'mos'; mysql@lists.mysql.com
Objet : RE: MySQL 5.0 creates 100,000 Window handles

Hi,

Try [EMAIL PROTECTED] instead.

Bye.
Geof.

-Message d'origine-
De : mos [mailto:[EMAIL PROTECTED] 
Envoyé : vendredi 25 mai 2007 06:41
À : mysql@lists.mysql.com
Objet : MySQL 5.0 creates 100,000 Window handles

I noticed if my program executes a lot of Select statements, Windows XP 
will slow down when the program completes. I did some investigating and 
mysqld-nt.exe has close to 100,000 handles created when my program ends 
(shown in Task Manager and SysInternals Process Explorer). As each Select 
statement is executed, 2 handles are created. These handles will stay 
allocated until the MySQL server is stopped (stopping my program won't free 
up the handles). Windows of course runs slow with this many handles
allocated.

1) What is MySQL using the handles for?
2) Is there any way to stop MySQL from consuming so many Windows resources?
3) Is it like this on Linux?
4) How does a MySQL server handle millions of queries on a large table? 
Will it hit an upper limit of Handles that it can allocate?

Note: it does not appear to allocate more handles if the query is found in 
the query cache.

Mike 

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Re: high physical and buffer reads.

2007-06-02 Thread Geoffroy Cogniaux

Hi,



I never heard about an Oracle's counter named HIGH PHYISCAL. But if you
think about Scattered Read / Sequential Read /  buffer busy waits,

You can begin by reading this article:

http://dev.mysql.com/tech-resources/articles/mysql-query-cache.html



Geoffroy.


2007/6/2, Baron Schwartz [EMAIL PROTECTED]:


Hi,

Ananda Kumar wrote:
 Hi All,
 Is it possible for us to know if a sql is doing HIGH PHYISCAL or BUFFER
 READS in mysql, something we can do in ORACLE.
 Just wanted to know, for the sake of sql tuning.

I am no Oracle expert, so I'm not sure I know exactly what you
mean.  However,
MySQL does expose many status counters and variables.  Learning them all
and
learning how to make sense of them is a chore.  Measuring them effectively
is
also somewhat tedious.  I wrote MySQL Query Profiler to help me with
this.  As
far as I know it is the only tool of its kind.  You can get it from
http://sourceforge.net/projects/mysqltoolkit/

Baron

--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:
http://lists.mysql.com/[EMAIL PROTECTED]




RE: mysql old 4.* query fails on 5.*

2007-06-02 Thread Geoffroy Cogniaux
Hi,

See 5.0 JOIN syntax at http://dev.mysql.com/doc/refman/5.0/fr/join.html

Example:
SELECT table1.* FROM table1
LEFT JOIN table2 ON table1.id=table2.id

Regards,
Geoffroy

-Message d'origine-
De : Gmail User [mailto:[EMAIL PROTECTED] 
Envoyé : samedi 2 juin 2007 22:55
À : mysql@lists.mysql.com
Objet : Re: mysql old 4.* query fails on 5.*

 Anyone know whats wrong here?

Try as 

...
From (Klienter AS K, Tid As Td, Personal AS P)
JOIN Uppdrag AS U ON K.Klient_ID = U.Klient_ID
...

or 

...
From Tid As Td, Personal AS P, Klienter AS K
JOIN Uppdrag AS U ON K.Klient_ID = U.Klient_ID
...

This is the problem I had in one of my queries. The join is on the last
table on the left side. Either use parentheses or put K table last.

HTH,

Ed


-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: MySQL 5.0 creates 100,000 Window handles

2007-06-02 Thread Geoffroy Cogniaux

You left in extreme tests. I would like to be able to answer that if it were
so problematic, it would have gone up for a long time in a list of the known
limits. But I do not know.

It seems that if you're really falling in this extreme case (many many very
small different queries in a big cache), regularly flushing query cache
appears to be a good maintenance plan.

It let me want to test what happens with a very big InnoDB table used in a
complex transaction such as deep update.

In any case, if you want to change your simple XP workstation into requested
server, it will be heroic.

Geoffroy

-Message d'origine-
De : mos [mailto:[EMAIL PROTECTED] 
Envoyé : samedi 2 juin 2007 23:49
À : mysql@lists.mysql.com
Cc : [EMAIL PROTECTED]
Objet : RE: MySQL 5.0 creates 100,000 Window handles

Geoffroy,
   Thanks for the reply. Of course by now I figured out what's causing 
the large # of handles being allocated by Windoze XP.  It's the query 
cache. Each query that gets added to the query cache uses 2 handles. If the 
query cache is large, say 150MB, as the cache fills up more and more 
handles are allocated. I was able to create a test program and generated 
hundreds of thousands of simple queries that returned a small # of rows 
from 1 large table (there were no cache hits because I wanted to fill the 
query cache up with as many unique queries as possible). I let it run 
overnight and in the morning Windows had allocated over 600,000 handles! It 
would have been more but the query cache was full  by then. If I were to 
increase the query_cache_size then I'm sure I could get XP to allocate over 
1 million handles. Flushing the query cache of course releases the handles. 
I'm not sure in the same thing happens in Linux or not.  Does it? Should I 
be worried running MySQL on an XP box that has to run 24/7?

Mike

At 02:20 PM 6/2/2007, Geoffroy Cogniaux wrote:
Hi,

In fact, this is a good question. Therefore, I gonna try to answer.

  1) What is MySQL using the handles for?
Handles are nothing and all on Windows. Nothing because they are only
pointers to hidden internal struct. And all, because, Handle are everywhere
if you try to develop Win32 App.
There are 3 kinds:
Users: Window, Cursor, Menu,...
GDI: all graphic objects such as Brush, Pen,...
Kernel: Access token (ACL), Console input, Event, File, Heap, Mutex, Pipe,
Process, Semaphore, Socket, Thread, Timer, ...
A database server is therefore a great consumer of kernel Handles.

  4) How does a MySQL server handle millions of queries on a large table?
  Will it hit an upper limit of Handles that it can allocate?
The per-process theoretical limit on kernel handles is 2^24. However,
handles are stored in the paged pool (kernel reserved memory), so the
actual
number of handles you can create is based on available memory. So, the
number of handles that you can create on 32-bit Windows is significantly
lower than 2^24. Example on Windows2000, max pool size is 300 MB (I don't
know on most recent windows versions).
But be sure MySQL server can handle millions of queries on a large table.

  2) Is there any way to stop MySQL from consuming so many Windows
  resources?
I investigated and can summarize with a simple test I made with a 5.0.37
compiled with all storage engines (and verified with a 4.1.21): MySQL
allocates 43000 handles. I recompiled it without InnoDB and BDB, and MySQL
allocates now 108 handles at startup. I'm not an indeep Innodb's
specialist neither BDB, but I know they have row locking mechanism in
difference of MyIsam but I'm sure that they are great consumers of Mutex.

  Windows of course runs slow with this many handles allocated.
Yes, not because of number of handles (logical resources) but because of
physical resources, especially RAM, and perhaps by-design in OS kernel.


Regards,
Geoffroy

-Message d'origine-
De : Geoffroy Cogniaux [mailto:[EMAIL PROTECTED]
Envoyé : dimanche 27 mai 2007 23:13
À : 'mos'; mysql@lists.mysql.com
Objet : RE: MySQL 5.0 creates 100,000 Window handles

Hi,

Try [EMAIL PROTECTED] instead.

Bye.
Geof.

-Message d'origine-
De : mos [mailto:[EMAIL PROTECTED]
Envoyé : vendredi 25 mai 2007 06:41
À : mysql@lists.mysql.com
Objet : MySQL 5.0 creates 100,000 Window handles

I noticed if my program executes a lot of Select statements, Windows XP
will slow down when the program completes. I did some investigating and
mysqld-nt.exe has close to 100,000 handles created when my program ends
(shown in Task Manager and SysInternals Process Explorer). As each Select
statement is executed, 2 handles are created. These handles will stay
allocated until the MySQL server is stopped (stopping my program won't free
up the handles). Windows of course runs slow with this many handles
allocated.

1) What is MySQL using the handles for?
2) Is there any way to stop MySQL from consuming so many Windows resources?
3) Is it like this on Linux?
4) How does a MySQL server handle millions of queries on a large table

Re: MySQL in multi-threaded environment

2007-05-31 Thread Geoffroy Cogniaux

Hi,
Assuming that you are using a multithread safe libmysql, I suggest you to
have a look at your error code first:
Error code 1064 suggests that you send a bad query to mysql, maybe your
pool-query isn't indeed MT safe, so manipulating this variable requires a
mutex.
Error code 1062 suggests that you try to insert duplicate key in your table:
remove this key ( bad suggestion ) or check that your pool doesn't send more
than one time the same query ( perhaps, the mutex on pool-query is enough )

Try to limit mutexes only where it's necessary and be aware where you're
locking and unlocking them.

Regards,
Geoffroy.

2007/5/30, Ace [EMAIL PROTECTED]:


Hi Friends,

I am facing problem while using MySQL in multi-threaded environment. I am
using C lang for developement.

I maintain MySQL connection pool between threads but with increase in
number
of requests, it started reporting
following errors -

==
Server Errors:
Error: 1064 :Parse error
Error: 1062 :Duplicate entry
==

Then I used mutexes around MySQL API calls and it worked. But seems use of
mutex impacts the performance, is this true?

Any other solution to this than mutex or any out-of-box solution that
might
have worked?

Thanks for your help!!!


--

Cheers,
Rajan



RE: mysql creating lots of processes (not threads, linux processes)

2007-05-30 Thread Geoffroy Cogniaux
Try to start it with mysql_safe instead or try to start mysqld manually
within a command prompt, without fork, to see what happen.
./mysqld --console --verbose --your_options

Can you at least connect to mysql with a remote client on this server or
not?

Have a look on this page about starting issues:
http://dev.mysql.com/doc/refman/5.0/en/unix-post-installation.html#starting-
server

Geoffroy

-Message d'origine-
De : [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] De la part de Quentin
Gouedard
Envoyé : mercredi 30 mai 2007 09:02
À : Scott Tanner
Cc : mysql@lists.mysql.com
Objet : Re: mysql creating lots of processes (not threads, linux processes)

Nope, I'm using 5.0.38 on Gentoo, built via emerge in the exact same manner.
Thanks for your answers guys.

On 5/30/07, Scott Tanner [EMAIL PROTECTED] wrote:

 Sounds like your not using threaded libraries. Was mysql built
 differently, or are you using a different RPM on this server?


 Scott



 On Tue, 2007-05-29 at 22:49 +0200, Quentin Gouedard wrote:
  No, I have just collectd+mrtg, but i don't even use them to monitor
 mysql.
  I launch mysql via /etc/init.d/mysql start , and the script is the exact
  same as on the other servers. Even just after startup there's already
 15-20
  processes created.
 
  On 5/29/07, Geoffroy Cogniaux [EMAIL PROTECTED] wrote:
  
   Hi,
  
 It looks like automatic start-up called by a monitoring process
 (Nagios,
   ...). Have you such tools on your servers ?
  
   Geoffroy
  
   -Message d'origine-
   De: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] De la part de
 Quentin
   Gouedard
   Envoyé: mardi 29 mai 2007 16:41
   À: mysql@lists.mysql.com
   Objet: mysql creating lots of processes (not threads, linux processes)
  
   Hi,
   So I use mysql as the DB for a large site (up to 1 concurrent
 users at
   peaks).
   I have a front server as a reverse proxy and multiple (7) backend
 machines
   serving the site.
   Each machine has data strictly similar in nature and quantity.
  
   On 6 of these machines, I have 1 single mysqld process (process in
 linux
   terms):
   # ps -ef | grep mysqld | wc -l
   2
  
   There are generally 5-8 threads (processes as mysql means it) running
 when
   i
   do a show processlist;
  
  
   Now, on one of those machines there are huge number of processes for
   mysql.
   # ps -ef | grep mysqld | wc -l
   34
   Running just ps shows for each of these processes:
   mysql25952 10073  0 16:25 ?00:00:02 /usr/sbin/mysqld
   --defaults-file=/etc/mysql/my.cnf --basedir=/usr
 --datadir=/var/lib/mysql
   --pid-file=/var/run/mysqld/mysqld.pid
 --socket=/var/run/mysqld/mysqld.sock
  
   This machine has no particular data, is doing nothing different than
 the
   others.
   The show processlist command also returns 5-8 processes.
  
   So where are these myqsld processes from ? There's like 20 at startup
   (instantly after launching mysql), but it keeps increasing, until i
   restart
   mysql or the server runs out of memory. I have compared the mysql
   configuration of this machine and the 6 other, variable by variable,
 and
   they are strictly identical.
   How come this server behaves differently ? What can I do to have the
   single-process behaviour on that machine too ?
  
   Thanks,
   Quentin
  
  
  





--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: mysql creating lots of processes (not threads, linux processes)

2007-05-29 Thread Geoffroy Cogniaux
Hi,

  It looks like automatic start-up called by a monitoring process (Nagios,
...). Have you such tools on your servers ?

Geoffroy

-Message d'origine-
De : [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] De la part de Quentin
Gouedard
Envoyé : mardi 29 mai 2007 16:41
À : mysql@lists.mysql.com
Objet : mysql creating lots of processes (not threads, linux processes)

Hi,
So I use mysql as the DB for a large site (up to 1 concurrent users at
peaks).
I have a front server as a reverse proxy and multiple (7) backend machines
serving the site.
Each machine has data strictly similar in nature and quantity.

On 6 of these machines, I have 1 single mysqld process (process in linux
terms):
# ps -ef | grep mysqld | wc -l
2

There are generally 5-8 threads (processes as mysql means it) running when i
do a show processlist;


Now, on one of those machines there are huge number of processes for mysql.
 # ps -ef | grep mysqld | wc -l
34
Running just ps shows for each of these processes:
mysql25952 10073  0 16:25 ?00:00:02 /usr/sbin/mysqld
--defaults-file=/etc/mysql/my.cnf --basedir=/usr --datadir=/var/lib/mysql
--pid-file=/var/run/mysqld/mysqld.pid --socket=/var/run/mysqld/mysqld.sock

This machine has no particular data, is doing nothing different than the
others.
The show processlist command also returns 5-8 processes.

So where are these myqsld processes from ? There's like 20 at startup
(instantly after launching mysql), but it keeps increasing, until i restart
mysql or the server runs out of memory. I have compared the mysql
configuration of this machine and the 6 other, variable by variable, and
they are strictly identical.
How come this server behaves differently ? What can I do to have the
single-process behaviour on that machine too ?

Thanks,
Quentin



--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: restore one database.

2007-05-28 Thread Geoffroy Cogniaux
Yes it is.
This is why I suggest dumping db per db.

Bye.

-Message d'origine-
De : Ananda Kumar [mailto:[EMAIL PROTECTED] 
Envoyé : lundi 28 mai 2007 10:41
À : Geoffroy Cogniaux
Cc : John Meyer; MySQL General
Objet : Re: restore one database.

Hi Geoffroy,
Very true, restore depends on the kind of backup we do. I was just wondering
if mysql has any option to restore just one database from the mysqldump
having all the database.

So, as of now mysql does not provided option to just restore just one
database from dump having all database..Right?

regards
anandkl


On 5/28/07, Geoffroy Cogniaux [EMAIL PROTECTED] wrote:

 Hi Ananda,

 Recovery strategy depends essentially of the backup strategy you choose. I
 think that if you want be able to proceed to a restoration, database per
 database, you should separate their backups. It can simply be done with
 this
 kind of script:

 for db in (`echo 'show databases;' | mysql -u user --password=pwd |
 grep
 -v ^Database$`); do  mysqldump -u user --password=pwd $db
 /mybackupdir/$db.bak ; done;

 If you want to have all in one file, use tar after your backup:
 cd /mybackupdir  tar -czf mybackup.tar.gz *.bak  rm -fr *.bak

 Best regards,
 Geof.

 -Message d'origine-
 De: Ananda Kumar [mailto:[EMAIL PROTECTED]
 Envoyé: dimanche 27 mai 2007 13:52
 À: John Meyer
 Cc: MySQL General
 Objet: Re: restore one database.

 Hi All,
 I think my question was not understood.
 All the database are important. Now that one of the database is accidently
 dropped, can i restore from that single database from my dump and use the
 bin log and recover till AS OF NOW.

 regards
 anandkl


 On 5/27/07, John Meyer [EMAIL PROTECTED] wrote:
 
  Ananda Kumar wrote:
   Hi Pelle,
   I dont have enough space on any other storage, so i was thinking if we
   would
   just restore one database from dump that would save lot  of time ,
  rather
   than restoring all the database.
  
   regards
   anandkl
  
  
  Well, if only one database is important enough to back up, then yes it
  will.  But if you have multiple databases that you are actively using
  then you'll need to back them all up.  You don't necessarily need to use
  an all databases dump, though.
 
 
  --
  The NCP Revue -- http://www.ncprevue.com/blog
 
 
  --
  MySQL General Mailing List
  For list archives: http://lists.mysql.com/mysql
  To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
 
 






--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



Re: CSV import

2007-05-28 Thread Geoffroy Cogniaux

Hi,
It can be done easily with PhpMyAdmin, but it is not in .net

2007/5/28, Sharique uddin Ahmed Farooqui [EMAIL PROTECTED]:


I want to import data from a CSV file in a  table. MySql admin doesn't
support import from CSV files.
Format of data is different from structure of table.

Is there any app/snippet   written for this task in .net , which I can
modify according to my need.

--
Sharique uddin Ahmed Farooqui
(C++/C# Developer, IT Consultant)
A revolution is about to begin.
A world is about to change.
And you and me are the initiator.



RE: error message

2007-05-27 Thread Geoffroy Cogniaux
Hi,
Could we show your procedure to try to help you ?

Geoffroy.

-Message d'origine-
De : Ananda Kumar [mailto:[EMAIL PROTECTED] 
Envoyé : vendredi 25 mai 2007 17:35
À : MySQL General
Objet : error message

Hi All,
We are using  5.0.40-enterprise-gpl-log version on our production db. If we
run a stored proc, we are getting the below error
Query OK, 0 rows affected, 1 warning (0.01 sec)

ERROR 1308 (42000): LEAVE with no matching label: SWL_return
mysql show warnings;
- //
+---+--+--+
| Level | Code | Message  |
+---+--+--+
| Error | 1308 | LEAVE with no matching label: SWL_return |
+---+--+--+

But on our qa with version 5.0.32-enterprise-gpl-log, the proc gets created
and exeuctes also.



Can you please help here.



regards

anandkl



--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: restore one database.

2007-05-27 Thread Geoffroy Cogniaux
Hi Ananda,

Recovery strategy depends essentially of the backup strategy you choose. I
think that if you want be able to proceed to a restoration, database per
database, you should separate their backups. It can simply be done with this
kind of script:

for db in (`echo 'show databases;' | mysql -u user --password=pwd | grep
-v ^Database$`); do  mysqldump -u user --password=pwd $db
/mybackupdir/$db.bak ; done;

If you want to have all in one file, use tar after your backup:
cd /mybackupdir  tar -czf mybackup.tar.gz *.bak  rm -fr *.bak

Best regards,
Geof.

-Message d'origine-
De : Ananda Kumar [mailto:[EMAIL PROTECTED] 
Envoyé : dimanche 27 mai 2007 13:52
À : John Meyer
Cc : MySQL General
Objet : Re: restore one database.

Hi All,
I think my question was not understood.
All the database are important. Now that one of the database is accidently
dropped, can i restore from that single database from my dump and use the
bin log and recover till AS OF NOW.

regards
anandkl


On 5/27/07, John Meyer [EMAIL PROTECTED] wrote:

 Ananda Kumar wrote:
  Hi Pelle,
  I dont have enough space on any other storage, so i was thinking if we
  would
  just restore one database from dump that would save lot  of time ,
 rather
  than restoring all the database.
 
  regards
  anandkl
 
 
 Well, if only one database is important enough to back up, then yes it
 will.  But if you have multiple databases that you are actively using
 then you'll need to back them all up.  You don't necessarily need to use
 an all databases dump, though.


 --
 The NCP Revue -- http://www.ncprevue.com/blog


 --
 MySQL General Mailing List
 For list archives: http://lists.mysql.com/mysql
 To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]





--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: restore one database.

2007-05-27 Thread Geoffroy Cogniaux
Hi Ananda,

Recovery strategy depends essentially of the backup strategy you choose. I
think that if you want be able to proceed to a restoration, database per
database, you should separate their backups. It can simply be done with this
kind of script:

for db in (`echo 'show databases;' | mysql -u user --password=pwd | grep
-v ^Database$`); do  mysqldump -u user --password=pwd $db
/mybackupdir/$db.bak ; done;

If you want to have all in one file, use tar after your backup:
cd /mybackupdir  tar -czf mybackup.tar.gz *.bak  rm -fr *.bak

Best regards,
Geof.

-Message d'origine-
De : Ananda Kumar [mailto:[EMAIL PROTECTED] 
Envoyé : dimanche 27 mai 2007 13:52
À : John Meyer
Cc : MySQL General
Objet : Re: restore one database.

Hi All,
I think my question was not understood.
All the database are important. Now that one of the database is accidently
dropped, can i restore from that single database from my dump and use the
bin log and recover till AS OF NOW.

regards
anandkl


On 5/27/07, John Meyer [EMAIL PROTECTED] wrote:

 Ananda Kumar wrote:
  Hi Pelle,
  I dont have enough space on any other storage, so i was thinking if we
  would
  just restore one database from dump that would save lot  of time ,
 rather
  than restoring all the database.
 
  regards
  anandkl
 
 
 Well, if only one database is important enough to back up, then yes it
 will.  But if you have multiple databases that you are actively using
 then you'll need to back them all up.  You don't necessarily need to use
 an all databases dump, though.


 --
 The NCP Revue -- http://www.ncprevue.com/blog


 --
 MySQL General Mailing List
 For list archives: http://lists.mysql.com/mysql
 To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]





--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]



RE: MySQL 5.0 creates 100,000 Window handles

2007-05-27 Thread Geoffroy Cogniaux
Hi,

Try [EMAIL PROTECTED] instead.

Bye.
Geof.

-Message d'origine-
De : mos [mailto:[EMAIL PROTECTED] 
Envoyé : vendredi 25 mai 2007 06:41
À : mysql@lists.mysql.com
Objet : MySQL 5.0 creates 100,000 Window handles

I noticed if my program executes a lot of Select statements, Windows XP 
will slow down when the program completes. I did some investigating and 
mysqld-nt.exe has close to 100,000 handles created when my program ends 
(shown in Task Manager and SysInternals Process Explorer). As each Select 
statement is executed, 2 handles are created. These handles will stay 
allocated until the MySQL server is stopped (stopping my program won't free 
up the handles). Windows of course runs slow with this many handles
allocated.

1) What is MySQL using the handles for?
2) Is there any way to stop MySQL from consuming so many Windows resources?
3) Is it like this on Linux?
4) How does a MySQL server handle millions of queries on a large table? 
Will it hit an upper limit of Handles that it can allocate?

Note: it does not appear to allocate more handles if the query is found in 
the query cache.

Mike 

-- 
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]




--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]