[ 
https://issues.apache.org/jira/browse/HADOOP-10705?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14036085#comment-14036085
 ] 

Colin Patrick McCabe commented on HADOOP-10705:
-----------------------------------------------

bq. Thanks Colin 's detail explanation. Try the in other Linux distro(CentOS 
6.4) ,the valgrind report about getpwuid_r go away, however, there is a report 
about uv__loop_init still exist, need to care about?

That one is real, and it's fixed by the patch I posted in HADOOP-10667.  
Basically, we weren't unregistering all the listeners before trying to shut 
down the event loop, which was causing the shutdown to fail.

> Fix namenode-rpc-unit  warning reported by memory leak check tool(valgrind)
> ---------------------------------------------------------------------------
>
>                 Key: HADOOP-10705
>                 URL: https://issues.apache.org/jira/browse/HADOOP-10705
>             Project: Hadoop Common
>          Issue Type: Sub-task
>    Affects Versions: HADOOP-10388
>            Reporter: Wenwu Peng
>         Attachments: vargrind.log
>
>
> Rum valgrind to check memory leak for namenode-rpc.unit.
> There are many warning, need to fix it.
> valgrind --tool=memcheck --leak-check=full --show-reachable=yes 
> ./namenode-rpc-unit 
> ==24085== 
> ==24085== HEAP SUMMARY:
> ==24085==     in use at exit: 1,216 bytes in 20 blocks
> ==24085==   total heap usage: 89 allocs, 69 frees, 12,222 bytes allocated
> ==24085== 
> ==24085== 128 bytes in 1 blocks are definitely lost in loss record 18 of 20
> ==24085==    at 0x4A069EE: malloc (vg_replace_malloc.c:270)
> ==24085==    by 0x4A06B62: realloc (vg_replace_malloc.c:662)
> ==24085==    by 0x4C2BD6A: uv__io_start (core.c:679)
> ==24085==    by 0x4C31868: uv_signal_init (signal.c:225)
> ==24085==    by 0x4C3000E: uv__loop_init (loop.c:136)
> ==24085==    by 0x4275EF: hrpc_reactor_create (reactor.c:183)
> ==24085==    by 0x426829: hrpc_messenger_create (messenger.c:83)
> ==24085==    by 0x42590F: main (namenode-rpc-unit.c:115)
> ==24085== 
> ==24085== LEAK SUMMARY:
> ==24085==    definitely lost: 128 bytes in 1 blocks
> ==24085==    indirectly lost: 0 bytes in 0 blocks
> ==24085==      possibly lost: 0 bytes in 0 blocks
> ==24085==    still reachable: 1,088 bytes in 19 blocks
> ==24085==         suppressed: 0 bytes in 0 blocks
> ==24085== Reachable blocks (those to which a pointer was found) are not shown.
> ==24085== To see them, rerun with: --leak-check=full --show-reachable=yes
> ==24085== 
> ==24085== For counts of detected and suppressed errors, rerun with: -v
> ==24085== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 6 from 6)



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Reply via email to