I'm using MySQL 4.0.12 on RedHat 7.1. I grabbed the binaries off mysql.com because I was having random table crashes like stated in the download section for gcc-2.96.*.
When I switched over to the binaries the crashes continued, but now they are getting pretty violent on one table. The table crashes so badly that I haven't been able to fix it with myisamchk -r/-o.
Here I paste an example of output from myisamchk:
# myisamchk -o myndir - recovering (with keycache) MyISAM-table 'myndir' Data records: 8221 Key 1 - Found wrong stored record at 114103880 Found link that points at 7080077846876182184 (outside data file) at 114104860 Found block with too small length at 114105416; Skipped Found block with too small length at 114105552; Skipped Found link that points at 31060663335674065 (outside data file) at 114105756 Found block with too small length at 114106036; Skipped Found link that points at -1447392181387475218 (outside data file) at 114106100 Found block with too small length at 114106192; Skipped Found link that points at -1432101689887775058 (outside data file) at 114106208 Found link that points at 770256467916358051 (outside data file) at 114106372 Found link that points at -6654559182842471983 (outside data file) at 114106376 Found link that points at -7674053380346937280 (outside data file) at 114107172 Found block with too small length at 114107440; Skipped Found link that points at 606174609524421989 (outside data file) at 114107520 Found block with too small length at 114107668; Skipped Found block with too small length at 114108196; Skipped Found link that points at -5398872103199542048 (outside data file) at 114108216 .... Found block with too small length at 114123420; Skipped Found link that points at -4538900753816510448 (outside data file) at 114123552 myisamchk: error: Not enough memory for blob at 114123616 myisamchk: Error writing file 'UNOPENED' (Errcode: 9) MyISAM-table 'myndir' is not fixed because of errors Try fixing it by using the --safe-recover (-o) or the --force (-f) option
I'm pretty sure this has nothing to do with disk space or memory, since this machine has enough of both (approx 20gb free hd, 3gb free of ram). Also, the datarecords it shows above, are more like 27000.
Can anyone help me fix this. Recovering the files from backup isn't fun to do 2-3 times a day. :/
Do you perhaps have an index bigger than 2GB on a filesystem with 2GB maximum size?
Guðmundur Bjarni Ólafsson wrote:
Hello,>
I'm using MySQL 4.0.12 on RedHat 7.1. I grabbed the binaries off mysql.com >because I was having random table crashes like stated in the download section >for gcc-2.96.*.>
When I switched over to the binaries the crashes continued, but now they are >getting pretty violent on one table. The table crashes so badly that I >haven't been able to fix it with myisamchk -r/-o.>
Here I paste an example of output from myisamchk:>
Guðmundur Bjarni Ólafsson 5 May 2003 16:44:58 [ permanent link ]
The datafile is about 300 megs, the index is alot smaller than that.
Sometimes I can fix the table, sometimes I cant.
regards GBO
On Monday 05 May 2003 13:28, gerald_clark wrote:> Do you perhaps have an index bigger than 2GB on a filesystem with 2GB> maximum size?>
Guðmundur Bjarni Ólafsson wrote:> >Hello,> >
I'm using MySQL 4.0.12 on RedHat 7.1. I grabbed the binaries off mysql.com> >because I was having random table crashes like stated in the download> > section for gcc-2.96.*.> >
When I switched over to the binaries the crashes continued, but now they> > are getting pretty violent on one table. The table crashes so badly that> > I haven't been able to fix it with myisamchk -r/-o.> >
Here I paste an example of output from myisamchk:> >
You do shut down the server before running myisamchk don't you?
Guðmundur Bjarni Ólafsson wrote:
The datafile is about 300 megs, the index is alot smaller than that.>
Sometimes I can fix the table, sometimes I cant.>
regards>GBO>
On Monday 05 May 2003 13:28, gerald_clark wrote:>
Do you perhaps have an index bigger than 2GB on a filesystem with 2GB>>maximum size?>>
Guðmundur Bjarni Ólafsson wrote:>>
Hello,>>>
I'm using MySQL 4.0.12 on RedHat 7.1. I grabbed the binaries off mysql.com>>>because I was having random table crashes like stated in the download>>>section for gcc-2.96.*.>>>
When I switched over to the binaries the crashes continued, but now they>>>are getting pretty violent on one table. The table crashes so badly that>>>I haven't been able to fix it with myisamchk -r/-o.>>>
Here I paste an example of output from myisamchk:>>>
I'm using MySQL 4.0.12 on RedHat 7.1. I grabbed the binaries off> >>> mysql.com because I was having random table crashes like stated in the> >>> download section for gcc-2.96.*.> >>>
When I switched over to the binaries the crashes continued, but now they> >>>are getting pretty violent on one table. The table crashes so badly that> >>>I haven't been able to fix it with myisamchk -r/-o.> >>>
Here I paste an example of output from myisamchk:> >>>
Datatal AB - Gauffin 7 October 2003 10:05:30 [ permanent link ]
Hello
Some one my tables have crashed (again). I had to repair them ASAP since the db is in my production enviroment. 1. How can I find the cause of the crashes?
2. I've created a my.cnf and added the following lines. Is the format correct? [mysqld] set-variable = myisam-recover=BACKUP,FORCE set-variable = myisam_sort_buffer_size=4M
I need to get this going without getting crashes every week, please help.
crashes occurs for some reasons like : - unclean tables ( tables are closed improprely), - hardware problems - something goes wrong in the system - know bugs Check that your tables are clean, before start using them.
It seems that you said the REPAIR doesn't occurs after you configuring related variables in the my.cnf file ? Because the daemon will not start when the my.cnf contains any errors (variable names), the format is to be correct.
Can you indicates what MySQL version you run and your OS version.
Regards Thierno6C - MySQL 4.0.15
----- Original Message ----- From: "Datatal AB - Gauffin, Jonas" <jonas@datatal.se> To: <mysql@lists.mysql.com> Sent: Tuesday, October 07, 2003 7:05 AM Subject: Table crashes
Hello
Some one my tables have crashed (again). I had to repair them ASAP since the db is in my production enviroment. 1. How can I find the cause of the crashes?
2. I've created a my.cnf and added the following lines. Is the format correct? [mysqld] set-variable = myisam-recover=BACKUP,FORCE set-variable = myisam_sort_buffer_size=4M
I need to get this going without getting crashes every week, please help.
MyODBC 3.51 is the only way that I access the database. (MyODBC is used in
a multithreaded server and in >InternetInformationServer (asp))>The server is coded in c++ and got a class that acts like a odbc pool (keeps connections open during the lifetime of the server >and handles them to threads that requests a db connection)
It may help to know more about the work of features (multithreaded server, 'odbc pool') you are using with MySQL. I have only one vague idea it above. I know that MySQL is multi-threaded (it may have many queries on the same table simultaneously).
Remarks that keeps connections open during the lifetime of the server and handles them to threads that requests a db connection SHOULDN'T be the culprit of table crash. In contrary this is a good thing; PHP, that i used, do the same job with the feature of PERSISTENT CONNECTIONS. Also the version of MyODBC (MyODBC 3.51 ) you are using is declared STABLE by MySQL , though MySQL says : 'some issues brought up appear to be application-related and independent of the ODBC driver or underlying database server.
How are tables getting unclean?
Has many ways like : - the mysql daemon server stops in a abnormal way , you can see it in the mysql error log file - a query fail to continue processing - a system problem
When are tables closed?
A table is opened for each concurrent acces, A table is closed when the connection using them is closed / lost
Regards Thierno6C - MySQL 4.0.15
----- Original Message ----- From: "Datatal AB - Gauffin, Jonas" <jonas@datatal.se> To: <mysql@lists.mysql.com> Sent: Tuesday, October 07, 2003 10:05 AM Subject: SV: Table crashes
Hello
crashes occurs for some reasons like :> - unclean tables ( tables are closed improprely),> - hardware problems> - something goes wrong in the system> - know bugs> Check that your tables are clean, before start using them.
MyODBC 3.51 is the only way that I access the database. (MyODBC is used in a multithreaded server and in InternetInformationServer (asp)) The server is coded in c++ and got a class that acts like a odbc pool (keeps connections open during the lifetime of the server and handles them to threads that requests a db connection)
How are tables getting unclean? When are tables closed? Should I disconnect/connect when a thread is handling the dbconnection back to the pool?
It seems that you said the REPAIR doesn't occurs after you> configuring related variables in the my.cnf file ? Because> the daemon will not start when the my.cnf contains any errors> (variable names), the format is to be correct.
No. I asked if my configuration were correct. I've just added it to my.cnf.
Can you indicates what MySQL version you run and your OS version.
Win2000 Mysql 4.0.12 (mysqld.exe) MyODBC was an old version (2.x), I've just installed 3.51.06
Being a system administrator, I periodically use various data recovery services from other software developers. From my point of view, the microsoft sql database repair utility: http://www.recoverytoolbox.com/microsoft_sql_server_database_repair_tool.html is very easy to use, it features intelligent data recovery algorithms that work under all supported software and hardware configurations.
If you would like to report an abuse of our service, such as a spam message, please . Åñëè Âû õîòèòå ïîæàëîâàòüñÿ íà ñîäåðæèìîå ýòîé ñòðàíèöû, ïîæàëóéñòà .
