[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 Dan Streetman changed: What|Removed |Added CC||ddstr...@ieee.org -- You are

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 Dan Streetman changed: What|Removed |Added Status|NEEDINFO|NEW --- Comment #18 from Dan

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #17 from Rafael David Tinoco --- Hello Eric, Sorry for taking so long in providing you feedback, I moved from one company to another and there is another engineer dealing with this case on the previous company (and I lost emails

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #16 from Eric Covener --- Hi Rafael -- out of curiosity did you ever find how fcntl() came to be used here? -- You are receiving this mail because: You are the assignee for the bug.

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #15 from Eric Covener --- [Tue May 08 08:26:42.418299 2018] [ldap:crit] [pid 68897:tid 78b1] (11)Resource deadlock avoided: [client 127.0.0.1:60404] LDAP cache lock failed -- You are receiving this mail

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #14 from Eric Covener --- In a local test I set: `Mutex fcntl:/tmp/run default` globally and added an assert to the two lock/unlock macros and got several aborts. The asserts with sysvsem never fired. I will

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #13 from Eric Covener --- > Another question that raises is, why to use fcntl as a backing mechanism for > the > LDAP locking ? If the lock was supposed to be guaranteed among different > nodes, > then backing

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #12 from Rafael David Tinoco --- Thread #19 7092 (Suspended : Container) kill() at syscall-template.S:84 0x7ff7e9911767 () at 0x7ff7e9cb7390 find_block_of_size() at apr_rmm.c:106

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #11 from Eric Covener --- Is fcntl being forced here somehow, or is it the default (apachectl -V|grep APR_USE) ? I wonder if an affected user could try a different mechanism, via e.g.

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #10 from Rafael David Tinoco --- Eric, Even after your fix... LDAP_CACHE_LOCK() is either missing a barrier or it is not enough for subsequent calls to APR with NULL locking (passed to APR_RMM_INIT).

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #9 from Rafael David Tinoco --- Just for a reference on someone that finds this bug. Distributions Fixes (Debian & Ubuntu): https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=814980#15

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #8 from Eric Covener --- I recently re-fixed an issue that causes the lock in mod_ldap to not be used. It will be in 2.4.31. It is a one-line change. http://svn.apache.org/viewvc?view=revision=1824811 -- You

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #7 from Rafael David Tinoco --- Max or Eric, Are there any changes in status of this bug ? I can corroborate to what Max said, and proposed, since I'm analysing a dump that was brought to me, exactly

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 Rafael David Tinoco changed: What|Removed |Added CC|

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 Guido W. changed: What|Removed |Added Status|NEW |NEEDINFO ---

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 Guido W. changed: What|Removed |Added Status|NEEDINFO|NEW -- You are

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #4 from Max Burke --- (In reply to Eric Covener from comment #3) > What kind of Require ldap-* do you use? I noticed the debian OP didn't list > that bit. I'm unfamiliar with what you mean by "Require ldap-*", but

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #3 from Eric Covener --- (In reply to Max Burke from comment #2) > I have not tried a later release in production, no, only because we have > been soaking the change to see if it improved reliability. > > I am

[Bug 60296] RMM list corruption in ldap module results in server hang

https://bz.apache.org/bugzilla/show_bug.cgi?id=60296 --- Comment #2 from Max Burke --- I have not tried a later release in production, no, only because we have been soaking the change to see if it improved reliability. I am looking at the mod_ldap changes between our current