The bypassing seems a little tricky, especially in a auto building production
env, in which one can not manually reload meta every time job failed.
The issue can severely damege the robustness of kylin cluster.
Is there any clue to fix the issue "It seems caused by the broadcaster doesn't
broadcast metadata to the job engine in time." ?
I found the snippet from source code :
CubeSegment segment = cube.getSegmentById(segmentID);
if (segment == null) {
logger.error("Failed to find {} in cube {}", segmentID, cube);
for (CubeSegment s : cube.getSegments()) {
logger.error(s.getName() + " with status " + s.getStatus());
}
throw new IllegalStateException();
}
At 2019-04-13 12:42:08, "Chao Long" <[email protected]> wrote:
It seems caused by the broadcaster doesn't broadcast metadata to the job engine
in time. There is a workaround you can try: reload the metadata and resume the
failed job.
There are two related issues:
https://issues.apache.org/jira/browse/KYLIN-3740
https://issues.apache.org/jira/browse/KYLIN-3752
On Fri, Apr 12, 2019 at 11:56 PM venkatesh Mora <[email protected]>
wrote:
I am running into the same. Can someone respond on this. I opened this a couple
of days ago.
On Apr 11, 2019, at 9:31 PM, bubugao0809 <[email protected]> wrote:
Hi all，
I am using kylin-2.2.0, and some of my job constantly failed at the third
stage of "Extract Fact Table Distinct Columns", with exception below:
java.lang.IllegalStateException
at
org.apache.kylin.engine.mr.steps.FactDistinctColumnsJob.run(FactDistinctColumnsJob.java:116)
at
org.apache.kylin.engine.mr.common.MapReduceExecutable.doWork(MapReduceExecutable.java:129)
at
org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:125)
at
org.apache.kylin.job.execution.DefaultChainedExecutable.doWork(DefaultChainedExecutable.java:64)
at
org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:125)
at
org.apache.kylin.job.impl.threadpool.DistributedScheduler$JobRunner.run(DistributedScheduler.java:171)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
result code:2
