[
https://issues.apache.org/jira/browse/HIVE-17684?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16190516#comment-16190516
]
Sahil Takiar commented on HIVE-17684:
-------------------------------------
Long term it would be good to have some type of global memory manager per Spark
executor that can determine the heap size is growing too big (it can do this by
measuring the amount of time spent doing GC and comparing it to amount of data
being reclaimed by GC - if lots of time is spent doing GC and not much memory
is being released, then the heap is probably full). It could then abort
specific Hive Tasks it thinks are using too much memory (the tasks can report
their memory usage to this global manager). This pattern should be better than
relying on threshold values for determining when to abort a task.
> HoS memory issues with MapJoinMemoryExhaustionHandler
> -----------------------------------------------------
>
> Key: HIVE-17684
> URL: https://issues.apache.org/jira/browse/HIVE-17684
> Project: Hive
> Issue Type: Bug
> Components: Spark
> Reporter: Sahil Takiar
> Assignee: Sahil Takiar
>
> We have seen a number of memory issues due the {{HashSinkOperator}} use of
> the {{MapJoinMemoryExhaustionHandler}}. This handler is meant to detect
> scenarios where the small table is taking too much space in memory, in which
> case a {{MapJoinMemoryExhaustionError}} is thrown.
> The configs to control this logic are:
> {{hive.mapjoin.localtask.max.memory.usage}} (default 0.90)
> {{hive.mapjoin.followby.gby.localtask.max.memory.usage}} (default 0.55)
> The handler works by using the {{MemoryMXBean}} and uses the following logic
> to estimate how much memory the {{HashMap}} is consuming:
> {{MemoryMXBean#getHeapMemoryUsage().getUsed() /
> MemoryMXBean#getHeapMemoryUsage().getMax()}}
> The issue is that {{MemoryMXBean#getHeapMemoryUsage().getUsed()}} can be
> inaccurate. The value returned by this method returns all reachable and
> unreachable memory on the heap, so there may be a bunch of garbage data, and
> the JVM just hasn't taken the time to reclaim it all. This can lead to
> intermittent failures of this check even though a simple GC would have
> reclaimed enough space for the process to continue working.
> We should re-think the usage of {{MapJoinMemoryExhaustionHandler}} for HoS.
> In Hive-on-MR this probably made sense to use because every Hive task was run
> in a dedicated container, so a Hive Task could assume it created most of the
> data on the heap. However, in Hive-on-Spark there can be multiple Hive Tasks
> running in a single executor, each doing different things.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
