Mauro, Thanks for that information. I've created a ticket to reflect helping users in this situation [1]. This is something that is very nuanced, so feel free to track the ticket's progress and discussion.
Are you using the bootstrap to build or are you running make manually? The bootstrap process attempts a parallel build, so that will cause memory usage to spike. I've had a number of memory constrained devices trigger the OOM killer or lock up if that was not present as a result of parallel builds. Let me know and we can account for this accordingly. Thanks [1] https://issues.apache.org/jira/browse/MINIFICPP-496 On Tue, May 15, 2018 at 8:58 AM, mauro <[email protected]> wrote: > Hi Mark, Andy, > > compiling in debug mode didn't work because when compiling the file > FlowController.cpp the device runs out of memory. Apparently it requires > more than 400Mb of memory which I don't have in my platform. (Please advise > if this is what you expect) > > Compiler crashes with error listed below: > > //----------------------------------------------------------------------------------------- > arm-poky-linux-gnueabi-g++: internal compiler error: Killed (program > cc1plus) > Please submit a full bug report, > with preprocessed source if appropriate. > See <http://gcc.gnu.org/bugs.html> for instructions. > libminifi/CMakeFiles/core-minifi.dir/build.make:1550: recipe for target > 'libminifi/CMakeFiles/core-minifi.dir/src/FlowController.cpp.o' failed > make[2]: *** [libminifi/CMakeFiles/core-minifi.dir/src/FlowController.cpp.o] > Error 4 > CMakeFiles/Makefile2:1988: recipe for target > 'libminifi/CMakeFiles/core-minifi.dir/all' failed > make[1]: *** [libminifi/CMakeFiles/core-minifi.dir/all] Error 2 > Makefile:149: recipe for target 'all' failed > make: *** [all] Error 2 > //-------------------------------------------------------------------------------------- > > I have been monitoring the free memory during the compilation process and it > goes from about 400Mb to few Mb until the process stops. > > However, going back to the original issue, I managed to run MiNiFi on the > device and connect to NiFi. > In order to do that I deleted the folder provenance_repository. The program > re-creates it again at the next restart and all works fine. There must have > been some corrupted data in that folder making the program crash. If I > restore old content program crashes again. > > Thanks for your time. > > Kind regards, > Mauro > > > > -- > Sent from: http://apache-nifi-users-list.2361937.n4.nabble.com/
