Reliability of _rowid over multiple database accesses
I'd like to use the _rowid capability of MySQL to improve the efficiency of a project, but I have concerns about the reliability of the number it returns. I understand that _rowid "may change over time depending on many factors," but I'd like to know whether this is true if the row is locked. In particular, can I perform a select on a row which returns a _rowid, modify the row data, and later do an UPDATE using the _rowid as index, without worrying that the _rowid has changed? Any help will be much appreciated. Guy L. Smith Lead Developer [EMAIL PROTECTED]
RE: aggravating failure symptoms and steps to remedy them...
Not sure I'll be helpful, but I ran into something similar yesterday. Not exactly sure how I fixed it, but this is what I believe. 1) You should create a virgin data directory per the standard procedures on the production machine (install_db and safe_mysqld). 2) Then, if you have existing data, copy over the existing data directory. > -Original Message- > From: Peter H. Burris [mailto:[EMAIL PROTECTED]] > Sent: Tuesday, June 19, 2001 1:49 PM > To: [EMAIL PROTECTED] > Subject: aggravating failure symptoms and steps to remedy them... > > > I'm a sysadmin who created mysql installs for users on linux x86 boxen; I > had to create self-contained, architecture-specific packages to comply > with our in-house means of handling software in my department. That took > three weeks to fine-tune and distribute and mysql seemed to run out of the > box, after some tuning on my build-host. > > Everything LOOKED good...until the user who called and said, "did you > change my password?" uh, no. "Why can't I get on Mysql, then?" uh, did > you do a ps -ef? "nothing happens. > > well, the host.some.domain.err says the following: > > 010619 16:39:30 mysqld started > 010619 16:39:30 /usr/local/mysql/libexec/mysqld: Can't find file: > './mysql/host.frm' (errno: 13) > 010619 16:39:30 mysqld ended > > searches on dejanews.com (or 'googlegroups, as they are known) > revealed someone had run into this and had been advised to run the > installdb script; I did that again, but now it doesn't do me much good, > so I try linking over /usr.../mysql/var/mysql to /usr/../,ysql/libexec, > but to no visible change in behavior, judging by stdout and the *.err log. > > What painfully simple, stupid thing am I missing? > > > - > 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 > - 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
Replication and redirection
OS: RedHat 6.2 on master and slave mySQL:3.23.32 on master and slave Problem: Replication is not working. Symptoms include: * Connecting OK. 'Show slave status' reports nominally. * Position on the log file increments correctly when updates occur on the master. * The tables do not get updated. The web hosting service (20after4.com = recommended - very helpful people there) has performed a replication between two of their servers, using the same master for my replication attempt as the master for theirs. Their replication worked. The main difference is that they did not use redirection and I did. Having the development database on this system, I do not want to replicate directly. So, when we synced I created a local database here (xxxprod), copied the snapshot of the production database to the xxxprod data directory here, then initiated replication. We have tried putting the replication commands in the my.cnf file and using the 'change master to' command, and get the same results. Here is the current my.cnf file. [mysqld] server-id=11 replicate-do-db=xxx replicate-rewrite-db=xxx->xxxprod Anyone have an idea what we may have done wrong? - 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: Upgrade and start-up failures
Tried both. 1st copied the data tree (wherein there is a host.frm, host.MYD, and host.MYI files). When that didn't work, I did the mysql_install_db, but that had no effect. I suspect that there is a ./configure option I missed that declares the default data file format to use, and that they were built differently. But I cannot yet find that in the docs. > -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On > Behalf Of Steve Brazill > Sent: Monday, June 18, 2001 6:11 PM > To: [EMAIL PROTECTED] > Cc: [EMAIL PROTECTED] > Subject: Re: Upgrade and start-up failures > > > It's looking for the "host" table (physical files) within the > 'mysql' data area > (not your own data area). > > Did you run "mysql_install_db" after installing the new version ? > (or copy the > "mysql" subdirectory from the old version) > > Guy Smith wrote: > > > I just upgraded from 3.23.27 to 3.23.32 (can't go higher than > that - have to > > sync development with production systems). > > > > I copied the /data directory to the new home. When I try to > start mySQL, it > > halts and the log file has the entry "Can't find file: > 'host.ISM' (errno: > > 2)". I checked the old mySQL /data directory, and there never > was any such > > file. I can still start the old server without error. > > > > Any suggestions are welcome. > > > > OS:RedHat Linux 6.2 > > DIST: Built from source > > > - > 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 > - 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
Upgrade and start-up failures
I just upgraded from 3.23.27 to 3.23.32 (can't go higher than that - have to sync development with production systems). I copied the /data directory to the new home. When I try to start mySQL, it halts and the log file has the entry "Can't find file: 'host.ISM' (errno: 2)". I checked the old mySQL /data directory, and there never was any such file. I can still start the old server without error. Any suggestions are welcome. OS:RedHat Linux 6.2 DIST: Built from source
Upgrade and start-up failures
I just upgraded from 3.23.27 to 3.23.32 (can't go higher than that - have to sync development with production systems). I copied the /data directory to the new home. When I try to start mySQL, it halts and the log file has the entry "Can't find file: 'host.ISM' (errno: 2)". I checked the old mySQL /data directory, and there never was any such file. I can still start the old server without error. Any suggestions are welcome. OS:RedHat Linux 6.2 DIST: Built from source
Re: MyODBC and #Delete / almost resolved
Thanks to Apollo and Lee for fast responses. Both recommended adding TimeStamp columns to the table. Lee also recommended adding an auto_increment column. What actually was the culprit was that in the suspect tables, I had a Primary Key on a TinyText field. I have not tried (a) converting the key type to a standard key or (b) changing the column type to char() or varchar(). However, removing the Primary Key made the table readable. However, I cannot edit the tables. Access claims all are unupdatable. I suspect this is due to not having a unique key for Access to force control through, but would like to have that confirmed. orig msg --- OS: NT 4.0, SP6 MDAC: 2.6 MyODBC: 2.50.36 MS-Access: 2000 (9.0.3821 SR-1) MySQL-OS: Linux, RedHat 6.2 MySql: 3.23.27-beta Symptom: Two of seven tables return all columns and rows as "#Deleted" (and wouldn't you know it, the two most important tables in the database). have been through the install process several times without a change in symptoms. Oh, the correct number of rows are returned - the data is just unusable. Any help is appreciated. - 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
MyODBC and #Delete
OS: NT 4.0, SP6 MDAC: 2.6 MyODBC: 2.50.36 MS-Access: 2000 (9.0.3821 SR-1) MySQL-OS: Linux, RedHat 6.2 MySql: 3.23.27-beta Symptom: Two of seven tables return all columns and rows as "#Deleted" (and wouldn't you know it, the two most important tables in the database). have been through the install process several times without a change in symptoms. Oh, the correct number of rows are returned - the data is just unusable. Any help is appreciated. - 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
