[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: Shell Fork Runaway Bug (ITS#1973)

There is a bug fix for this in HEAD.  It is suggested
that older versions of back-shell not be used in
threaded environments. 


At 11:46 AM 2002-07-22, anthm@cylynx.com wrote:
>I didnt submit the original bug I just was a victim of it.
>When you use the shell backend there is code in both the shell mod and the main
>process that works find for any number of days then unde some condition
>will start to spawn off zombies like mad and take the box down.
>I do remem
>The issue tracking system used to find a bunch of refs to it.
>2.0.18 was when I first encountered it and used a patch
>as well as some rearranging of my own to get something to work
>nothing formal.
>I tried 2.0.25 expecting the bug to be gone and got the same
>problem was what I was trying to say.
>I looked it up on google and found this little thread discussing it.
>----- Original Message -----
>From: "Kurt D. Zeilenga" <Kurt@OpenLDAP.org>
>To: <anthm@cylynx.com>
>Cc: <openldap-its@OpenLDAP.org>
>Sent: Monday, July 22, 2002 2:08 AM
>Subject: Re: Shell Fork Runaway Bug (ITS#1973)
>At 08:47 AM 2002-07-21, anthm@cylynx.com wrote:
>>Full_Name: Anthony Minessale
>>Version: newest
>>OS: solaris
>>URL: ftp://ftp.openldap.org/incoming/
>>Submission from: (NULL) (
>>There is a bug in the code for the shell module
>>that causes exec / fork runaway that will kill the box.
>>It was already acknoledged like a year ago and I patched an
>>earlier version,  Is this bug going to be removed from the production
>>release soon?  I got the newest one last week and was suprised it still
>>and I have'nt the time to mess with patching it every time so
>>I guess I'll wait paitently.  Just wondering in case it slipped through
>>the cracks.
>What's the ITS tracking number of your original bug report / patch
>submission?  I'm also not sure what you mean by "newest".  Please
>use version numbers (or dates if using CVS branch tags).