[ 
https://issues.apache.org/jira/browse/THRIFT-3848?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15312572#comment-15312572
 ] 

James E. King, III commented on THRIFT-3848:
--------------------------------------------

Found this while root causing the issue:

http://stackoverflow.com/questions/7662711/perl-forked-socket-server-stops-accepting-connections-when-a-client-disconnects

> As an implementer of a perl socket server, I do not want to have to remember 
> to ignore SIGCHLD for it to work properly
> ----------------------------------------------------------------------------------------------------------------------
>
>                 Key: THRIFT-3848
>                 URL: https://issues.apache.org/jira/browse/THRIFT-3848
>             Project: Thrift
>          Issue Type: Bug
>          Components: Perl - Library
>    Affects Versions: 0.8, 0.9, 0.9.1, 0.9.2, 0.9.3
>         Environment: Ubuntu 14.04, perl forking server, C++ client
>            Reporter: James E. King, III
>            Assignee: James E. King, III
>
> In a project I work on, we use a perl thrift server as a mock to simulate 
> something in production.  A C++ client connects, makes a call, and 
> disconnects.  The perl server is a forking server.
> I found that if I do not explicity ignore SIGCHLD, I can accept one 
> connection and process it, but when it disconnects, the subsequent accept 
> call is interrupted by a SIGCHLD and results in a useless handle that 
> can_read says cannot be read.  This puts Server.pm into an infinite 
> accept/read/fail loop.
> Question to those in the know... should perl's ForkingServer serve() 
> implementation explicitly set SIGCHLD to IGNORE to avoid this situation?  If 
> we do not, folks trying to use it might be as confounded as I was as to why 
> this is happening.  If we do, it might change the behavior of programs using 
> the perl forking server.
> My inclination is to document this behavior, that perl's ForkingServer will 
> set SIGCHLD to IGNORE for proper operation.
> Tagging [~djnym], [~jensg], [~nsuke] for comment.  I've observed this 
> behavior as far back as thrift-0.8.0.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to