Full_Name: Howard Chu
Version:
OS:
URL: ftp://ftp.openldap.org/incoming/
Submission from: (NULL) (84.127.144.82)
Submitted by: hyc
It's possible for a write txn to update a meta page while a reader is creating
its snapshot of the meta page. In most cases this situation is harmless, but if
the update raises the mm_last_pg field it may later cause an assert if the
reader still has the old value of mm_last_pg, but accesses a DB whose pages are
beyond the old value.
A fix has been tested and will be committed shortly.
Thanks Micheal.
-----Original Message-----
From: Michael Ströder [mailto:michael@stroeder.com]
Sent: Monday, July 01, 2013 1:49 PM
To: KUMAR, Madhan (Madhan); openldap-its(a)openldap.org
Subject: Re: (ITS#7634) openldap 2.4.24 hangs while attempting to shutdown
madhan.kumar(a)alcatel-lucent.com wrote:
> We have an issue of openldap 2.4.24 hanging. This is similar to the issue
> mentioned in the link:
> http://www.openldap.org/lists/openldap-technical/201208/msg00119.html
If you read the follow-up in the ML archive Howard says this was fixed in 2.4.25.
So could you please try out recent release 2.4.35?
Ciao, Michael.
madhan.kumar(a)alcatel-lucent.com wrote:
> We have an issue of openldap 2.4.24 hanging. This is similar to the issue
> mentioned in the link:
> http://www.openldap.org/lists/openldap-technical/201208/msg00119.html
If you read the follow-up in the ML archive Howard says this was fixed in 2.4.25.
So could you please try out recent release 2.4.35?
Ciao, Michael.
Full_Name: Madhan Kumar
Version: 2.4.24
OS: linux
URL: ftp://ftp.openldap.org/incoming/
Submission from: (NULL) (135.245.192.4)
We have an issue of openldap 2.4.24 hanging. This is similar to the issue
mentioned in the link:
http://www.openldap.org/lists/openldap-technical/201208/msg00119.html
When I checked the openldap logs I could see that the listeners are shown as
busy continuously and an automatic internal shutdown is initiated to recover the
ldap process. Unfortunately in our case the automatic internal shutdown is stuck
and hence the issue.
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: epoll: listen=7 active_threads=0 tvp=zero
daemon: epoll: listen=8 busy
daemon: shutdown requested and initiated.
daemon: closing 7
daemon: closing 8
We need to get an confirmation that this issue is being fixed in openldap 2.4.25
so that we can upgrade our openldap version. Please confirm.