Re: drop table is written to bin-log, load table is NOT - why????
Hello Artem, thanks for your answer. For selective replication I would check startup options --replicate-do-* and --replicate-wild-*. I'm using replicate-do-db on the client but the problem is, that the master sends the sql-statements of ALL databases to the repl-client and the replicate-do-db-command only tells the client which sql-statements it should carry out and which statements it should drop without notice. In the actual case only 5% of the traffic is caused by the database which needs to be replicated so we would have to pay for 95% useless traffic if we wouldn't use daisy-chain Replication to filter the traffic. . Also you may need log-slave-update option on if you have daisy-chain replication. I'm using this option too but it seems to have no effect doing a "load table ... from master". Greetings from Germany Lutz -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
drop table is written to bin-log, load table is NOT - why????
Hello, I'm wondering why on a replication client (MySQL 4.0.23) a "drop table XYZ" is written to the mysql binlog while the following "load table XYZ from master" is not. This missing binlog-entry in the first repl. client causes a second replication client, which is replicating the first client, (sounds silly but this construction is necessary to filter the entries of a single database out of dozen hosted on the master and to replicate the logentries for this database only) to go out of sync cause it still holds the old values while the first client now is up to date. Help would be appreciated Lutz Maibach EasyCom GmbH -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: How to bring up local replication / socket-error
Hello Gleb, thanks for your hint - I will try this workaround to get the replication up and running. Is there a point on your TODO-List that sounds like 'splitting binary logs per database to enable replicating a single database instead of transfering Gigs of traffic and using only few bytes of it'? :-) It also would be very useful to add a feature on the master to define which databases/tables should or shouldn't be logged into the bin-log. Greetings from Germany Lutz - Original Message - From: "Gleb Paharenko" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Tuesday, November 09, 2004 3:49 PM Subject: Re: How to bring up local replication / socket-error Hello. See: http://bugs.mysql.com/bug.php?id=3310 "Lutz Maibach" <[EMAIL PROTECTED]> wrote: Hi, I'm trying to set up a replication client on the same server as the master is running. Master is port 3306 and its socket /mymaster/mysql.sock. I set up the repl-client to connect to localhost (tried the ip-address too) on port 3306. When I start the replication I get the following error: Error: 'Can't connect to local MySQL server through socket '/tmp/mysql.sock' (2)' errno: 2002 The question is how to tell the replication client that the right socket is located in /mysqlmaster. I tried to set a symbolic link from /tmp/mysql.sock to mysqlmaster.sock but only got a fatal error 1236: "Could not open logfile.." Help would be appreciated Greetz from Germany Lutz Maibach P.S: I only try to get up this strange looking constellation cause I have to replicate a single database from the master to a customers server. The customer may not have a look at the other databases running on the master, but with normal replication he would get all the data into his relaylog. So I want to replicate the customers database to the local replication and the the customers server to replicate this local replication where only his data is stored in the binlog. -- For technical support contracts, goto https://order.mysql.com/?ref=ensita This email is sponsored by Ensita.NET http://www.ensita.net/ __ ___ ___ __ / |/ /_ __/ __/ __ \/ /Gleb Paharenko / /|_/ / // /\ \/ /_/ / /__ [EMAIL PROTECTED] /_/ /_/\_, /___/\___\_\___/ MySQL AB / Ensita.NET <___/ www.mysql.com -- 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]
How to bring up local replication / socket-error
Hi, I'm trying to set up a replication client on the same server as the master is running. Master is port 3306 and its socket /mymaster/mysql.sock. I set up the repl-client to connect to localhost (tried the ip-address too) on port 3306. When I start the replication I get the following error: Error: 'Can't connect to local MySQL server through socket '/tmp/mysql.sock' (2)' errno: 2002 The question is how to tell the replication client that the right socket is located in /mysqlmaster. I tried to set a symbolic link from /tmp/mysql.sock to mysqlmaster.sock but only got a fatal error 1236: "Could not open logfile.." Help would be appreciated Greetz from Germany Lutz Maibach P.S: I only try to get up this strange looking constellation cause I have to replicate a single database from the master to a customers server. The customer may not have a look at the other databases running on the master, but with normal replication he would get all the data into his relaylog. So I want to replicate the customers database to the local replication and the the customers server to replicate this local replication where only his data is stored in the binlog. -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
mysql 4.1: how to use old-passwords in my.cnf
Hi, i'm testing 4.1.05, executed the mysql_fix_privilege_tables script but have to use "old clients (3.51.06)" to connect to mysql. For this reason I inserted the old-passwords parameter into my.cnf (I'm calling several mysql-daemons with different versions using a single script, so I don't wanted to use --old-passwords as a commandline-parameter) but it seems to be useless. [mysqld] old-passwords What do I need to write into my.cnf to get the old clients connecting again? Greetings Lutz Maibach -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
mysqladmin extended-status: useless statistics when running server for a while?
Hello, I tried to optimize our server according to the hints given in the mysql manual and some articles I read. To find out whether our system needs tuning I performed what's written in the manual section 4.6.8.3: If Key_reads is big, then your key_buffer_size variable is probably too small. The cache miss rate can be calculated with Key_reads/Key_read_request The rate was way to low after running the server for app. 80 days, so I doubled the key-cache and started again, controlling the ratio every 48 hours. Rate seemed to be much better (1/200 instead of 1/20 before) but from one day to the other the result dropped down on 1/4. Knowing that it can't be I controlled the absolute values and figured out, that the Key_reads-variable had an overflow reaching 4Billion. My conclusion was that the value is saved as an unsigned-int which is far to small for a long running statistic if it should be usefull longer than a couple of days. Now my question, especially for the Mysql-developers is, if it is possible to store this values as float or another type to make the statistics usable for a longer time. Our server isn't really a big one, but the overflow happened after running the server only for 11 days. Greetings from Germany Lutz Maibach -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
relay_log_space_limit-entry kills replication
Hello, trying to save hd-space we configured a 4.0.17-Replicationclient in my.cnf as shown below: max_relay_log_size= 250M relay_log_space_limit = 1000M After restarting Mysql a new relay_log was created correctly every 250M, so we thought everything was working well. The disaster occured when the relay_log_space_limit of 1.000M was reached cause replication stopped right at this moment. The error message leaded us into a wrong direction, so we didn't recognize immediately that the "space_limit" was the reason of our problems. Below you see the error message we found in the mysql.error-log: = Slave: Could not parse relay log event entry. The possible reasons are: the master's binary log is corrupted (you can check this by running 'mysqlbinlog' on the binary log), the slave's relay log is corrupted (you can check this by r unning 'mysqlbinlog' on the relay log), a network problem, or a bug in the master's or slave's MySQL code. If you want to che ck the master's binary log or slave's relay log, you will be able to know their names by issuing 'SHOW SLAVE STATUS' on this slave. Error_code: 0 = Finally we commented the entries out, re-synchronised the databases manually and restarted replication. The ultimate question for us now is "why didn't mysql delete old relay-logs to get space to continue logging, like it is described in the Mysql documentattion"? Is this a bug and, if so, is it solved in the actual version of mysql 4.0? Greetings Lutz Maibach EasyCom GmbH -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: How to minimize Master-Slave Traffic during replication <- Activating slave_compressed_protocol question
Hello Sasha, thanks for your advise, but I'm not shure whether I understand how to find out if the master is offering (and using) the compressed protocol. When I execute the strings-command, I get can see the string "slave_compressed_protocol" two times, which I expected running 4.0.18, but if I do a mysqld --help to see the standard values for the variables, I see "slave_compressed_protocol FALSE". Do I have to start mysqld with an entry like "slave_compressed_protocol=ON" in my.cnf or something like this to get the master running with a compressed protocol? Greetings Lutz Maibach - Original Message - From: "Sasha Pachev" <[EMAIL PROTECTED]> To: "Lutz Maibach" <[EMAIL PROTECTED]> Cc: "mysql Liste" <[EMAIL PROTECTED]> Sent: Thursday, March 11, 2004 11:32 PM Subject: Re: How to minimize Master-Slave Traffic during replication? > Lutz Maibach wrote: > > Hi, > > > > we got a "little problem" with a master-slave replication (both running > > MySQL 4.0.18) eating up our complete bandwidth. The slave is connected > > via a 2MBit-SDSL-Line which is also used to connect our Office-PCs with > > the internet so I get complaints about the "slow connection" which is > > caused by the huge amount of master-slave traffic through this line. > > Most of the replication traffic is caused by tables, which are only > > created once, used for further selects and then deleted. The tables > > can't be created as temporary tables (that's what our programmers are > > telling me - I'm only the Admin and don't know whether thei're right or > > wrong) but are completely useless for replication. > > Use SET SQL_LOG_BIN=0 on the connection (process privilege needed) to turn off > binary logging for the queries that do not need to be replicated to the slave. > SET SQL_LOG_BIN=1 to turn logging back on. > > > > > Does anyone use the slave_compressend_protocol - variable successfully > > and can tell me, where I can see, whether the slave recognized this > > switch and do I have to set this switch on the server too? > > It's a fairly new feature, but it should be very safe - it just enables the use > of some very well field tested code. You should set it only on the slave - the > slave will tell the master that it wants to use compression. To see if the > daemon has it, strings mysqld | grep slave_compression_protocol > > -- > Sasha Pachev > Create online surveys at http://www.surveyz.com/ > -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
How to minimize Master-Slave Traffic during replication?
Hi, we got a "little problem" with a master-slave replication (both running MySQL 4.0.18) eating up our complete bandwidth. The slave is connected via a 2MBit-SDSL-Line which is also used to connect our Office-PCs with the internet so I get complaints about the "slow connection" which is caused by the huge amount of master-slave traffic through this line. Most of the replication traffic is caused by tables, which are only created once, used for further selects and then deleted. The tables can't be created as temporary tables (that's what our programmers are telling me - I'm only the Admin and don't know whether thei're right or wrong) but are completely useless for replication. My first attempt to reduce traffic between master and slave was to set replicate-wild-ignore-table for the tables described above. Ok, after reading the MySQL-Manual several times I discovered that this was useless cause the insert, update and delete statements are transfered from master to slave before they are tested whether to be executed on the slave - it would be nice if this could be set in the master-config for every slave, so the useless sql-statements would'nt have to be transmitted. Now I want to test the "slave_compressed_protocol=1" to reduce traffic (meanwhile I estimated that app. 90% of the traffic between master and slave is "useless" for replication). First I want to test it out with another replication-slave where I set the variable and restarted the slave. Unfortunately I don't see any slave_compressed_protocol-Variable when I do a "mysqladmin variables" so I don't know whether the variable is recognized or not. Does anyone use the slave_compressend_protocol - variable successfully and can tell me, where I can see, whether the slave recognized this switch and do I have to set this switch on the server too? Greetings Lutz Maibach -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: Inserting One new User = 2 entries in mysql.user when using MySQL 4.0.17?!
Hi,
finally I solved the problem by myself, yippiee :-)
During the upgrade from 3.x to 4.x I didn't create the indizes for the
user, host und db-tables (I didn't use the mysql-fix-privilege-tables
skript cause I did not whether there was one at all). The missing
indizes caused MySQL to create a new user every time I granted rights
to an user, whether it was existing or not.
Greetings
Lutz
- Original Message -
From: "Lutz Maibach" <[EMAIL PROTECTED]>
To: "mysql Liste" <[EMAIL PROTECTED]>
Sent: Monday, January 26, 2004 3:01 PM
Subject: Inserting One new User = 2 entries in mysql.user when using
MySQL 4.0.17?!
> Hello,
>
> I'm using the same sqls to append a new user to the MySQL-DB since 2
> years.After upgrading from 3.23.57 to 4.0.17 I added a
replication-user
> with the same script that always worked right. When controlling the
new
> entry I was really surprised that I found two entries for the new user
> instead of one.
>
> To get rid of this "uninvited guest" I studied the manual again and
> found out that in MySQL 4.0.x a replication user now longer got the
> file_priv but the replication_slave_priv instead. So I deleted the two
> user-entries, changed the script to the one shown below
and.got
> again 2 entries for the new user which are different in the privileges
> they own.
>
> Here are the SQL-Statements I used:
> insert into user (host,user,password) values
> ('%','repltest2',password('test1234'));
> flush privileges;
> grant super, reload, replication slave on *.* to repltest2@'%'
> identified by 'test1234';
> flush privileges;
>
> ...and here is what I see in the mysql.user-table:
>
+--+-+--+-+-+---
> --+-+-+---+-+-
--
> +--+---++-+---
--
> ---++--++---+-
--
> ---+--+-+--+--
--
> --++-+--+---+-
--
> --+-+
> | Host | User| Password | Select_priv | Insert_priv |
> Update_priv | Delete_priv | Create_priv | Drop_priv | Reload_priv |
> Shutdown_priv | Process_priv | File_priv | Grant_priv |
References_priv
> | Index_priv | Alter_priv | Show_db_priv | Super_priv |
> Create_tmp_table_priv | Lock_tables_priv | Execute_priv |
> Repl_slave_priv | Repl_client_priv | ssl_type | ssl_cipher |
x509_issuer
> | x509_subject | max_questions | max_updates | max_connections |
>
+--+-+--+-+-+---
> --+-+-+---+-+-
--
> +--+---++-+---
--
> ---++--++---+-
--
> ---+--+-+--+--
--
> --++-+--+---+-
--
> --+-+
> | %| repltest2 | 38fb843c55682a6f | N | N | N
> | N | N | N | Y | N
|
> N| N | N | N | N |
N
> | N| Y | N | N
|
> N| Y | N| |
> | | | 0 | 0 |
> 0 |
> | %| repltest2 | 38fb843c55682a6f | N | N | N
> | N | N | N | N | N
|
> N| N | N | N | N |
N
> | N| N | N | N
|
> N| N | N| |
> | | | 0 | 0 |
> 0 |
>
+--+-+--+-+-+---
> --+-+-+---+-+-
--
> +--+---++-+---
--
> ---++--++---+-
--
> ---+--+---------+--+--
--
> --++-+--+---+-
--
> --+-+
>
> "User 1" got the right privileges while "User 2" haven't got a single
> privilege.
>
> Anyone knows what's going wrong?
>
> Greetings from Germany
>
> Lutz Maibach
>
>
> --
> 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]
Inserting One new User = 2 entries in mysql.user when using MySQL 4.0.17?!
Hello,
I'm using the same sqls to append a new user to the MySQL-DB since 2
years.After upgrading from 3.23.57 to 4.0.17 I added a replication-user
with the same script that always worked right. When controlling the new
entry I was really surprised that I found two entries for the new user
instead of one.
To get rid of this "uninvited guest" I studied the manual again and
found out that in MySQL 4.0.x a replication user now longer got the
file_priv but the replication_slave_priv instead. So I deleted the two
user-entries, changed the script to the one shown below and.got
again 2 entries for the new user which are different in the privileges
they own.
Here are the SQL-Statements I used:
insert into user (host,user,password) values
('%','repltest2',password('test1234'));
flush privileges;
grant super, reload, replication slave on *.* to repltest2@'%'
identified by 'test1234';
flush privileges;
...and here is what I see in the mysql.user-table:
+--+-+--+-+-+---
--+-+-+---+-+---
+--+---++-+-
---++--++---+---
---+--+-+--+
--++-+--+---+---
--+-+
| Host | User| Password | Select_priv | Insert_priv |
Update_priv | Delete_priv | Create_priv | Drop_priv | Reload_priv |
Shutdown_priv | Process_priv | File_priv | Grant_priv | References_priv
| Index_priv | Alter_priv | Show_db_priv | Super_priv |
Create_tmp_table_priv | Lock_tables_priv | Execute_priv |
Repl_slave_priv | Repl_client_priv | ssl_type | ssl_cipher | x509_issuer
| x509_subject | max_questions | max_updates | max_connections |
+--+-+--+-+-+---
--+-+-+---+-+---
+--+---++-+-
---++--++---+---
---+--+-+--+
--++-+--+---+---
--+-+
| %| repltest2 | 38fb843c55682a6f | N | N | N
| N | N | N | Y | N |
N| N | N | N | N | N
| N| Y | N | N|
N| Y | N| |
| | | 0 | 0 |
0 |
| %| repltest2 | 38fb843c55682a6f | N | N | N
| N | N | N | N | N |
N| N | N | N | N | N
| N| N | N | N|
N| N | N| |
| | | 0 | 0 |
0 |
+--+-+--+-+-+---
--+-+-+---+-+---
+--+---++-+-
---++--++---+---
---+--+-+--+
--++-+--+---+---
--+-+
"User 1" got the right privileges while "User 2" haven't got a single
privilege.
Anyone knows what's going wrong?
Greetings from Germany
Lutz Maibach
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Unexpected shutdown of MySQL4.0.17
Hi, yesterday I set up a Mysql4.0.17-Classic on an new Server with Redhat 9. Today MySQL shut down unexpectedly. The Logfile showed the following last entry: 040107 10:36:298124 Connect [EMAIL PROTECTED] on 8124 Query SHOW VARIABLES LIKE 'pid_file' ..and the error-log: = 040107 10:36:29 /usr/local/mysqlmaster/bin/mysqld: Normal shutdown A mysqld process already exists at Mit Jan 7 10:36:29 CET 2004 040107 10:36:30 /usr/local/mysqlmaster/bin/mysqld: Shutdown Complete 040107 10:36:30 mysqld ended I'm using a start-script for mysql since more than one year which was functioning very well with all 3.23-x Versions. It starts Mysql with the following line: mybase="/usr/local/mysqlmaster" myuser="mysqlec" $mybase/bin/safe_mysqld --defaults-file=$mybase/data/my.cnf --user=$myus er & I'm wondering why Mysql didn't restart after the unexpected shutdown and if something is wrong in the new acutal safe_mysqld? Greetings Lutz Maibach -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: Starting mysqld within SuSE 8.0
Hi Markus,
> can someone tell me how to start mysqld at boot on a SuSE 8.0 system?
> Starting it manually with safe_mysqld runs fine. But I didn't find a
way
> in YaST to start the daemon on boot, and manually adding it to the
> /etc/init.d/boot.d/ lead me into a crash.
The easiest way would be using Yast, then choosing System,
Runlevel-Editor. There you find Runlevel-Properties
("Runlevel-Eigenschaften" if you choose German Language for Yast). In
the upcoming window for the services ("Dienste") you can set markers, at
which runlevels mysql should start.
For more infos you should look here ->
http://sdb.suse.de/de/sdb/html/start_foo80.html
Greetings from Hessen/Germany
Lutz Maibach
EasyCom GmbH
-
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
Re: hi
Hi, > Can naybody tell me how can i get the last 10 inserted record in the table > in MySql. Is MySql uses rowid like Oracle. Yep, you can use _rowid and a user variable to get what you want: select @startrow:=count(*)-10 from mytable select * from mytable where _rowid>@startrow Greetings Lutz Maibach - 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
Rename Table in Replication failed / command missing in slave log
Hi, today I noticed a strange behaviour in MySQL 3.23.49a-Replication I can't explain. A perl-script which controls whether master- and replication-tables are equal showed me, that a master-table was missing on both slaves (two backups are better than one :-) ) When I controlled the master update-log I saw that the master-table was renamed with the following command: ALTER TABLE ad_429_t RENAME adprj_7; This command was missing in the slave-updatelogs. The sql-statements before and right after the missing one were present and no error was written down in the mysql-error-log. Trying to find out whether the Rename-Statement works right or not I did some tests with renaming a table but all actions were performed on the slaves too. Has someone experienced the same? I'm now a bit unsure whether the replication is working right. Greetings from Germany Lutz Maibach - 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
Error using 'select...into outfile'
Hi, I'm trying to do the following: select * into outfile '/tmp/data.txt' fields terminated by';' from customer Although the mysql-user has all privileges (except the grant priv) on the database and the file_priv in the user table I get an 'Access denied for user..' when executing the select from a remote odbc-tool (QTODBC). Trying this select with the same user and password on the unix console /tmp/data.txt is created and filled. I don't know why the behaviour for the same user is different when I'm sending the same selects via odbc instead of sending it from the unix console. Does anyone know a solution for this problem? Greetings from Germany Lutz - 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
Aborted connection using myodbc
Hi, we're using myodbc2.50.36 with W2k and W98 to connect to our mysql-server (Linux Kernel 2.2.18 - MySQL Version 3.23.38) in the internet. When we send or receive larger amounts of data (we're using Visual Foxpro 7.0 as frontend) we get the "aborted connections"-Message in the mysql errorlog very often. In my.cnf the wait_timeout is set to 1800 seconds but the connection is aborted much earlier. Here's a piece of the most recent logfile: 020215 16:22:22 Aborted connection 22987 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:25:27 Aborted connection 22991 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:26:02 Aborted connection 22999 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error writing communication packets) 020215 16:28:58 Aborted connection 23574 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:32:38 Aborted connection 24381 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:34:41 Aborted connection 24877 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:34:53 Aborted connection 24970 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error reading communication packets) 020215 16:43:23 Aborted connection 25296 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got an error writing communication packets) 020215 16:48:47 Aborted connection 22968 to db: 'dummy' user: 'dummyusr' host: `my.server.com' (Got timeout reading communication packets) Does anybody know what to do to stop this aborted connections? It's nearly impossible to send or receive larger amounts of data when manipulating tables using ODBC. Greetings from Germany Lutz Maibach - 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
Maximum parallel Users on MySQL
Hi, is there a limit of concurrent connections or threads in MySQL? What do I have to consider when many users (500 - 1000) want to use MySQL at the same time? Greetings from Germany Lutz Maibach EasyCom GmbH
Re: Visual Foxpro 6
Hi Christopher, >>I wish to use MySQL as a backend to a Visual Foxpro 6 client application, >>has anyone got any experience of this ? Does it work ? We're using MySQL as backend for our VFP6-Application. It's working with the MyODBC-Driver but you have to take care about memo-fields and boolean fields. We're no longer use boolean cause it isn't supported by MySQL. So we changed to the good old fashioned binary value (tinyint in MySQL) with 0 and 1. Using Memofields is a bit more tricky cause you have to use text or blob in MySQL to store their content. MySQL has no problems storing it but when you're trying to get the data back to VFP you can't use a passthrough select statement. If you do so VFP handles the content of the blob field as an object, not as text. For this reason you have to use a view. In the view definition you can change the type for the blob field from object to memo so it will be displayed right in VFP. Oh, I almost forgot the nastiest point up to now, the view definition: When you create a view with VFP the MyODBC-Driver loads down the complete table content before showing the dialog where you can choose the fields from the table structure. If you got a tabelle with a million rows (like we do) you can spend much time in drinking coffee and walking around before you get this dialog. So create the views on an empty table and fill them after creating the view. Greetings from cloudy Germany Lutz Maibach EasyCom GmbH - 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
mysql.pm or dbi, that's the question
Hi, we're using mysql.pm to connect to mysql. It's more comfortable than using dbi directly so we don't want to rewrite our sources for using dbi directly if there's no good reason to do so. When I examined the mysql.pm I found the sentence "You are strongly encouraged to implement new code with the DBI Drivers". Is this a "must do" or only a "can do"? Will there be a dramatic change in DBI so Mysql.pm wouldn't be able to use it anymore? Questions, so many questions. Anyone out there who got some reassuring answers for me? :-) Greetings Lutz Maibach
Using Mysql.pm with a different port
Hello, we're using mysql.pm to connect to mysql from perl. After installing a second Mysql-Server we need to connect to the new installation using another port. Unfortunately we did not find out how to do this up to now. Setting mysql_tcp_port or mysql_unix_port in the perl-script doesn't work. Trying to set the port for the DBI-Connect in Mysql.pm failed too. We can't use DBI instead of mysql.pm cause we're using the mysql.pm- Query subroutine in almost every script (>100). Greetings from Germany Lutz Maibach
