I found that GC was the main problem in my circumstance, each of my node throws the GC exception: ---------------------------------------------------------------------------------------------------------------- Exception in thread "shmem-worker-#175%null%" java.lang.OutOfMemoryError: GC overhead limit exceeded [05-Apr-2016 19:05:26][ERROR][shmem-worker-#176%null%][TcpCommunicationSpi] Runtime error caught during grid runnable execution: ShmemWorker [endpoint=IpcSharedMemoryClientEndpoint [inSpace=IpcSharedMemorySpace [opSize=262144, shmemPtr=139664015134784, shmemId=2883604, semId=2392067, closed=true, isReader=true, writerPid=11421, readerPid=11230, tokFileName=/opt/apache-ignite-1.5.0.final-src/work/ipc/shmem/a3bcd536-31b5-47f6-b248-80f5a43e50dc-11230/gg-shmem-space-46-11421-262144, closed=true], outSpace=IpcSharedMemorySpace [opSize=262144, shmemPtr=139663894958144, shmemId=2916373, semId=2424836, closed=true, isReader=false, writerPid=11230, readerPid=11421, tokFileName=/opt/apache-ignite-1.5.0.final-src/work/ipc/shmem/a3bcd536-31b5-47f6-b248-80f5a43e50dc-11230/gg-shmem-space-47-11421-262144, closed=true], checkIn=true, checkOut=true]] java.lang.OutOfMemoryError: GC overhead limit exceeded ----------------------------------------------------------------------------------------------------------------
-- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Why-the-client-and-server-behaves-like-that-tp3935p3937.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
