strange regression in ldap support between 5.3 and 5.4

strange regression in ldap support between 5.3 and 5.4

[Guillaume Rousse]

Hello list.

I have a user reporting a strange regression after updating from linux 
mandriva 2009.0 (shipping autofs 5.0.3 + all patches until 
fix-couple-of-memory-leaks.patch) to mandriva 2009.1 (shipping autofs 
5.4 + all patches until clear-rpc-client-on-lookup-fail.patch): 
apparently, keys defined in indirect maps are mounted directly on the 
mount point, instead of a subdir:

https://qa.mandriva.com/show_bug.cgi?id=50631

The debug output is apparently not consistent with mount output:
mounted /home/lsc/luizd
vs
sol.lsc.ufsc.br:/export/lsc/home/luizd on /home/lsc type nfs

Curiously, file based maps works OK. But, the only ldap-related changes 
I see in the applied patches set (5.0.3 with all patches is basically 
5.0.4) is a memory leak fix, and a build-time detection fix.

Does someones see something else ?

-- 
BOFH excuse #323:

Your processor has processed too many instructions.  Turn it off 
immediately, do not type any commands!!

Re: strange regression in ldap support between 5.3 and 5.4

[Ian Kent]

Guillaume Rousse wrote:
> Hello list.
> 
> I have a user reporting a strange regression after updating from linux 
> mandriva 2009.0 (shipping autofs 5.0.3 + all patches until 
> fix-couple-of-memory-leaks.patch) to mandriva 2009.1 (shipping autofs 
> 5.4 + all patches until clear-rpc-client-on-lookup-fail.patch): 
> apparently, keys defined in indirect maps are mounted directly on the 
> mount point, instead of a subdir:
> 
> https://qa.mandriva.com/show_bug.cgi?id=50631
> 
> The debug output is apparently not consistent with mount output:
> mounted /home/lsc/luizd
> vs
> sol.lsc.ufsc.br:/export/lsc/home/luizd on /home/lsc type nfs
> 
> Curiously, file based maps works OK. But, the only ldap-related changes 
> I see in the applied patches set (5.0.3 with all patches is basically 
> 5.0.4) is a memory leak fix, and a build-time detection fix.
> 
> Does someones see something else ?

Looks like the trailing "/" on the autofs mount point might be causing
some confusion:

dn: cn=/home/lsc/,nisMapName=auto.master,ou=autofs,ou=nss,dc=lsc,dc=ufsc,dc=
 br
cn: /home/lsc/
nisMapName: /home/lsc/
objectClass: nisObject
objectClass: top
nisMapEntry: ldap:lua.lsc.ufsc.br:nisMapName=auto.home,ou=autofs,ou=nss,dc=l
 sc,dc=ufsc,dc=br


I'd be interested to know if this is the case.

Re: strange regression in ldap support between 5.3 and 5.4

[Ian Kent]

Ian Kent wrote:
> Guillaume Rousse wrote:
>> Hello list.
>>
>> I have a user reporting a strange regression after updating from linux 
>> mandriva 2009.0 (shipping autofs 5.0.3 + all patches until 
>> fix-couple-of-memory-leaks.patch) to mandriva 2009.1 (shipping autofs 
>> 5.4 + all patches until clear-rpc-client-on-lookup-fail.patch): 
>> apparently, keys defined in indirect maps are mounted directly on the 
>> mount point, instead of a subdir:
>>
>> https://qa.mandriva.com/show_bug.cgi?id=50631
>>
>> The debug output is apparently not consistent with mount output:
>> mounted /home/lsc/luizd
>> vs
>> sol.lsc.ufsc.br:/export/lsc/home/luizd on /home/lsc type nfs
>>
>> Curiously, file based maps works OK. But, the only ldap-related changes 
>> I see in the applied patches set (5.0.3 with all patches is basically 
>> 5.0.4) is a memory leak fix, and a build-time detection fix.
>>
>> Does someones see something else ?
> 
> Looks like the trailing "/" on the autofs mount point might be causing
> some confusion:
> 
> dn: cn=/home/lsc/,nisMapName=auto.master,ou=autofs,ou=nss,dc=lsc,dc=ufsc,dc=
>  br
> cn: /home/lsc/
> nisMapName: /home/lsc/

And shouldn't nisMapName be auto.master?

> objectClass: nisObject
> objectClass: top
> nisMapEntry: ldap:lua.lsc.ufsc.br:nisMapName=auto.home,ou=autofs,ou=nss,dc=l
>  sc,dc=ufsc,dc=br
> 
> 
> I'd be interested to know if this is the case.
> 
> _______________________________________________
> autofs mailing list
> autofs@linux.kernel.org
> http://linux.kernel.org/mailman/listinfo/autofs

Re: strange regression in ldap support between 5.3 and 5.4

[Guillaume Rousse]

Ian Kent a écrit :
> I'd be interested to know if this is the case.
You were right :)

-- 
BOFH excuse #361:

Communist revolutionaries taking over the server room and demanding all 
the computers in the building or they shoot the sysadmin. Poor misguided 
fools.