hive-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "bd2019us (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HIVE-21555) [SECURITY]newCachedThreadPool() has higher risk in causing OutOfMemoryError
Date Mon, 01 Apr 2019 02:21:00 GMT

     [ https://issues.apache.org/jira/browse/HIVE-21555?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

bd2019us updated HIVE-21555:
----------------------------
    Description: 
Location: ql/src/java/org/apache/hadoop/hive/ql/processors/LlapCacheResourceProcessor.java:136

In the program, since the number of instances (instances = llapRegistryService.getInstances().getAll();)
is not known ahead of time, it has a high risk in causing OutOfMemoryError when too many threads
are created. Therefore, to ensure security, a fixed number of thread pool should be used,
which can be freely configured and also enables running multiple instances concurrently but
is free from the above error.

PR: https://github.com/apache/hive/pull/585

  was:
Location: ql/src/java/org/apache/hadoop/hive/ql/processors/LlapCacheResourceProcessor.java:136

In the program, since the number of instances (instances = llapRegistryService.getInstances().getAll();)
is not known ahead of time, it has a high risk in causing OutOfMemoryError when too many threads
are created. Therefore, to ensure security, a fixed number of thread pool should be used,
which can be freely configured and also enables running multiple instances concurrently but
is free from the above error.


> [SECURITY]newCachedThreadPool() has higher risk in causing OutOfMemoryError
> ---------------------------------------------------------------------------
>
>                 Key: HIVE-21555
>                 URL: https://issues.apache.org/jira/browse/HIVE-21555
>             Project: Hive
>          Issue Type: Bug
>            Reporter: bd2019us
>            Assignee: bd2019us
>            Priority: Major
>              Labels: patch
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> Location: ql/src/java/org/apache/hadoop/hive/ql/processors/LlapCacheResourceProcessor.java:136
> In the program, since the number of instances (instances = llapRegistryService.getInstances().getAll();)
is not known ahead of time, it has a high risk in causing OutOfMemoryError when too many threads
are created. Therefore, to ensure security, a fixed number of thread pool should be used,
which can be freely configured and also enables running multiple instances concurrently but
is free from the above error.
> PR: https://github.com/apache/hive/pull/585



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Mime
View raw message