You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Experienced this when shrinking the hbz01XmlClobs.tar.bz2 test set to one, while there are two threads to compute the data. When calling closeStream(), a NullPointer Exception appears. This stops the ETL process finishing correctly, and thus no resource (aka document) is serialized to the disk.
A workaround is to catch the Exception.
However, as this is a problem can occur everywhere, it would be nicer to somehow control this at the modules in metafacture-core, probably at the ObjectPipeDecoupler. Also, to say the least, it's rather unexpected to use more threads than documents, it may happen.
The text was updated successfully, but these errors were encountered:
If there are more threads than resources to be computed "closeStream()"
interrupts the computation of all resources. Catching the Exception is
enough to make it finishing correctly.
See #1030.
()
Experienced this when shrinking the hbz01XmlClobs.tar.bz2 test set to one, while there are two threads to compute the data. When calling
closeStream()
, a NullPointer Exception appears. This stops the ETL process finishing correctly, and thus no resource (aka document) is serialized to the disk.A workaround is to catch the Exception.
However, as this is a problem can occur everywhere, it would be nicer to somehow control this at the modules in metafacture-core, probably at the ObjectPipeDecoupler. Also, to say the least, it's rather unexpected to use more threads than documents, it may happen.
The text was updated successfully, but these errors were encountered: