vBulletin Search Engine Optimization
| |||||||
| Register | FAQ | Members List | Calendar | Search | Today's Posts | Mark Forums Read |
| ||||
| I have a Python application which seems to lose its connection to the database exactly once per week. It's not an idle time issue; there's a transaction every few seconds, continuously. Is there a 1-week timeout somewhere in MySQL or MySQLdb? John Nagle |
| |||
| John Nagle schreef: > I have a Python application which seems to lose its connection > to the database exactly once per week. It's not an idle time issue; > there's a transaction every few seconds, continuously. Is there > a 1-week timeout somewhere in MySQL or MySQLdb? > > John Nagle does this happen when your logfiles rotate ? (just a wild guess...) -- Luuk |
| |||
| Luuk wrote: > John Nagle schreef: >> I have a Python application which seems to lose its connection >> to the database exactly once per week. It's not an idle time issue; >> there's a transaction every few seconds, continuously. Is there >> a 1-week timeout somewhere in MySQL or MySQLdb? >> >> John Nagle > > does this happen when your logfiles rotate ? (just a wild guess...) It happens every Sunday morning, when the first new transaction comes in after 5 AM or so. Unfortunately, I don't have the error message from Python, because I'm logging errors in the database, not a log file. This has happened for three weeks in a row now, and it's the only time the app crashes. I run a big batch update from another process every Sunday from 0030 to 0330 or so. But that's loading some tables that shouldn't affect the Python application at all. The update creates some new tables and does a LOAD DATA LOCAL INFILE. At the end of the update, the tables are renamed. But the tables are completely different from the ones used by the app that crashes. The Linux server and mysqld instance have been up without a crash for months. It's not a database engine restart. John Nagle |
| |||
| Hi John, John Nagle <nagle@animats.com> wrote: > Luuk wrote: >>> I have a Python application which seems to lose its connection >>> to the database exactly once per week. It's not an idle time issue; >>> there's a transaction every few seconds, continuously. Is there >>> a 1-week timeout somewhere in MySQL or MySQLdb? > It happens every Sunday morning, when the first new transaction > comes in after 5 AM or so. > > The Linux server and mysqld instance have been up without a crash for > months. It's not a database engine restart. I guess you checked this in the MySQL error log? No other messages for this time? You should add log_warnings=2 to my.cnf. This will enable some additional messages in the MySQL error log. http://dev.mysql.com/doc/refman/5.0/...d_log-warnings XL -- Axel Schwenke, MySQL Support Engineer, Sun Microsystems GmbH MySQL User Manual: http://dev.mysql.com/doc/refman/5.0/en/ MySQL User Forums: http://forums.mysql.com/ |
| |||
| Axel Schwenke wrote: > Hi John, > > John Nagle <nagle@animats.com> wrote: >> Luuk wrote: > >>>> I have a Python application which seems to lose its connection >>>> to the database exactly once per week. It's not an idle time issue; >>>> there's a transaction every few seconds, continuously. Is there >>>> a 1-week timeout somewhere in MySQL or MySQLdb? > >> It happens every Sunday morning, when the first new transaction >> comes in after 5 AM or so. >> >> The Linux server and mysqld instance have been up without a crash for >> months. It's not a database engine restart. > > I guess you checked this in the MySQL error log? No other messages > for this time? Nothing. This is the MySQL daemon log for all of 2008: 080120 22:34:17 mysqld started 080120 22:34:17 InnoDB: Started; log sequence number 1 192541792 080120 22:34:17 [Note] /usr/libexec/mysqld: ready for connections. Version: '5.0.27' socket: '/var/lib/mysql/mysql.sock' port: 3306 Source distribution > You should add log_warnings=2 to my.cnf. This will > enable some additional messages in the MySQL error log. OK, let me see how I can do that without making the Plesk control panel system unhappy. This is on a dedicated server, but it's maintained in a way that doesn't allow too many changes to the configuration. Is it "log_warnings" or "log-warnings"? John Nagle |
| ||||
| John Nagle schreef: > Axel Schwenke wrote: >> Hi John, >> >> John Nagle <nagle@animats.com> wrote: >>> Luuk wrote: >> >>>>> I have a Python application which seems to lose its connection >>>>> to the database exactly once per week. It's not an idle time issue; >>>>> there's a transaction every few seconds, continuously. Is there >>>>> a 1-week timeout somewhere in MySQL or MySQLdb? >> >>> It happens every Sunday morning, when the first new transaction >>> comes in after 5 AM or so. >>> >>> The Linux server and mysqld instance have been up without a crash >>> for >>> months. It's not a database engine restart. >> >> I guess you checked this in the MySQL error log? No other messages >> for this time? > > Nothing. This is the MySQL daemon log for all of 2008: > > 080120 22:34:17 mysqld started > 080120 22:34:17 InnoDB: Started; log sequence number 1 192541792 > 080120 22:34:17 [Note] /usr/libexec/mysqld: ready for connections. > Version: '5.0.27' socket: '/var/lib/mysql/mysql.sock' port: 3306 > Source distribution > > > You should add log_warnings=2 to my.cnf. This will >> enable some additional messages in the MySQL error log. > > OK, let me see how I can do that without making the Plesk control > panel system unhappy. This is on a dedicated server, but it's maintained > in a way that doesn't allow too many changes to the configuration. > > Is it "log_warnings" or "log-warnings"? > > John Nagle http://dev.mysql.com/doc/refman/5.0/en/error-log.html (so its, 'log_warnings' in your config file..) -- Luuk |