[MonetDB-users] MALException:opt.remap:no implementation present
Our application includes multiple services that independently connect to the database. When each service connects it reads the database metadata to discover the column names, many services may start at about the same time.
We are getting the following error:
java.sql.SQLException: TypeException:user.s7172_2[262]:'io.stdout' undefined in: _416:any MALException:opt.remap:no implementation present at nl.cwi.monetdb.jdbc.MonetConnection$ResponseList.executeQuery(MonetConnection.java:2058) at nl.cwi.monetdb.jdbc.MonetConnection$ResponseList.processQuery(MonetConnection.java:1808) at nl.cwi.monetdb.jdbc.MonetStatement.internalExecute(MonetStatement.java:371) at nl.cwi.monetdb.jdbc.MonetStatement.execute(MonetStatement.java:340) at nl.cwi.monetdb.jdbc.MonetStatement.executeQuery(MonetStatement.java:404) at nl.cwi.monetdb.jdbc.MonetDatabaseMetaData.getTypeInfo(MonetDatabaseMetaData.java:2454)
# check status of database
/monetdb status name state uptime crashes BUSINESS running 1d 1h 40m 38s 3 2008-04-22 11:23:59
# try a query using mclient running on the same machine as the database
./mclient -dBUSINESS -umonetdb -Pmonetdb -lsql sql>select * from tables; !TypeException:user.s5301_3[130]:'io.stdout' undefined in: _193:any sql>select * from schemas; !TypeException:user.s5302_3[40]:'io.stdout' undefined in: _63:any sql>select * from columns; !TypeException:user.s5303_3[146]:'io.stdout' undefined in: _210:any sql>\q;
When we stop and restart the merovingian demon, the problem goes away, but may reoccur the next time we try and start multiple services in short succession.
Environment
MonetDB MSG merovingian[12454]: starting database 'BUSINESS', up min/avg/max: 121/43165/92925, crash average: 0.00 0.30 0.10 (7-4=3) MSG BUSINESS[12459]: executing '/monetdb/bin/mserver5' for database 'BUSINESS' MSG BUSINESS[12459]: # MonetDB server v5.5.0, based on kernel v1.23.0 MSG BUSINESS[12459]: # Serving database 'BUSINESS' MSG BUSINESS[12459]: # Compiled for x86_64-unknown-linux-gnu/64bit with 64bit OIDs dynamically linked
JDBC Driver 1.7 (Canephora_p2 20080418 based on MCL v1.3)
On 23-04-2008 13:30:18 -0400, McKennirey.Matthew wrote:
Our application includes multiple services that independently connect to the database. When each service connects it reads the database metadata to discover the column names, many services may start at about the same time.
We are getting the following error:
java.sql.SQLException: TypeException:user.s7172_2[262]:'io.stdout' undefined in: _416:any MALException:opt.remap:no implementation present
Something goes horribly wrong deep down in the engine.
When we stop and restart the merovingian demon, the problem goes away, but may reoccur the next time we try and start multiple services in short succession.
Isn't stopping and starting the BUSINESS database enough to get it running again? Essentially, restarting merovingian is the same, but if you'd have more databases it could be a bit disruptive, I guess.
MSG merovingian[12454]: starting database 'BUSINESS', up min/avg/max: 121/43165/92925, crash average: 0.00 0.30 0.10 (7-4=3) MSG BUSINESS[12459]: executing '/monetdb/bin/mserver5' for database 'BUSINESS' MSG BUSINESS[12459]: # MonetDB server v5.5.0, based on kernel v1.23.0 MSG BUSINESS[12459]: # Serving database 'BUSINESS' MSG BUSINESS[12459]: # Compiled for x86_64-unknown-linux-gnu/64bit with 64bit OIDs dynamically linked
JDBC Driver 1.7 (Canephora_p2 20080418 based on MCL v1.3)
JDBC and Merovingian unrelated, my bet is concurrency is your enemy here.
Thanks, yes we had ruled out JDBC and Merovingian on this.
We have only seen this on the recent build we took (2008-04-21 nightly=current) of v5.5.0 We did not experience it on earlier versions.
On Wednesday 23 April 2008 14:11:29 Fabian Groffen wrote:
On 23-04-2008 13:30:18 -0400, McKennirey.Matthew wrote:
Our application includes multiple services that independently connect to the database. When each service connects it reads the database metadata to discover the column names, many services may start at about the same time.
We are getting the following error:
java.sql.SQLException: TypeException:user.s7172_2[262]:'io.stdout' undefined in: _416:any MALException:opt.remap:no implementation present
Something goes horribly wrong deep down in the engine.
When we stop and restart the merovingian demon, the problem goes away, but may reoccur the next time we try and start multiple services in short succession.
Isn't stopping and starting the BUSINESS database enough to get it running again? Essentially, restarting merovingian is the same, but if you'd have more databases it could be a bit disruptive, I guess.
MSG merovingian[12454]: starting database 'BUSINESS', up min/avg/max: 121/43165/92925, crash average: 0.00 0.30 0.10 (7-4=3) MSG BUSINESS[12459]: executing '/monetdb/bin/mserver5' for database 'BUSINESS' MSG BUSINESS[12459]: # MonetDB server v5.5.0, based on kernel v1.23.0 MSG BUSINESS[12459]: # Serving database 'BUSINESS' MSG BUSINESS[12459]: # Compiled for x86_64-unknown-linux-gnu/64bit with 64bit OIDs dynamically linked
JDBC Driver 1.7 (Canephora_p2 20080418 based on MCL v1.3)
JDBC and Merovingian unrelated, my bet is concurrency is your enemy here.
On 23-04-2008 15:40:32 -0400, McKennirey.Matthew wrote:
Thanks, yes we had ruled out JDBC and Merovingian on this.
We have only seen this on the recent build we took (2008-04-21 nightly=current) of v5.5.0 We did not experience it on earlier versions.
Feels related to a recent checkin of Martin: http://article.gmane.org/gmane.comp.db.monetdb.scm/5819 (and the other files modified in that commit)
Fabian Groffen wrote:
On 23-04-2008 15:40:32 -0400, McKennirey.Matthew wrote:
Thanks, yes we had ruled out JDBC and Merovingian on this.
We have only seen this on the recent build we took (2008-04-21 nightly=current) of v5.5.0 We did not experience it on earlier versions.
Feels related to a recent checkin of Martin: http://article.gmane.org/gmane.comp.db.monetdb.scm/5819 (and the other files modified in that commit)
those changes were committed on 23/4 not on 21/4 and not related to the error message.
If you perform an cvs up of these changes it is adviced to recompile MonetDB5 from scratch, removing its build. This due to possible errors introduced by incomplete reference checking and old code files.
I have to see the query (+schema) to understand what part of the query space is touched and not handled correctly. Don't hesitate to sent it.
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javao... _______________________________________________ MonetDB-users mailing list MonetDB-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-users
participants (3)
-
Fabian Groffen -
Martin Kersten -
McKennirey.Matthew