openldap-bugs

openldap-bugs@openldap.org

1 participants
21001 discussions

(ITS#6070) slapo-rwm(5) be_entry_get() incorrect behavior when no DN rewriting occurs
by ando＠sys-net.it 20 Apr '09

20 Apr '09

Full_Name: Pierangelo Masarati Version: HEAD/re24 OS: irrelevant URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (131.175.154.146) Submitted by: ando To reproduce, stack a slapo-rwm with no specific instructions on top of a provider right before the syncprov overlay. A fix is coming.

1 0

Re: (ITS#6055) Samba4 need 'name' implementation like AD (RDN-Name)
by michael＠stroeder.com 18 Apr '09

18 Apr '09

Pierangelo Masarati wrote: > michael(a)stroeder.com wrote: >> (I've stumbled across ITS#5704. Isn't that already something like this?) > > not exactly, as sets do not allow a placeholder for the naming > attribute; maybe something like "this/entryRDN.ava{0}.value" if it > existed... But given that slapd already enforces the characteristic attribute (e.g. 'cn' for 'user' on Samba4) to be part of the entry IMHO this is not needed. Provided the order of the checks are right. So this would be: constraint_attribute name set "this/cn" restrict="ldap:///dc=example,dc=com??sub?(objectClass=user)" constraint_attribute name set "this/ou" restrict="ldap:///dc=example,dc=com??sub?(objectClass=organizationalUnit)" Since uniqueness of 'cn' or 'ou' for one-level search for the above cases is already enforced wouldn't this be sufficient? Maybe I miss something though. Ciao, Michael.

1 0

Re: (ITS#6055) Samba4 need 'name' implementation like AD (RDN-Name)
by ando＠sys-net.it 18 Apr '09

18 Apr '09

michael(a)stroeder.com wrote: > ando(a)sys-net.it wrote: >> ----- abartlet(a)samba.org wrote: >>> Active Directory always presents an attribute 'name' that is always >>> equal to the relative distinguished name. AD allows only one RDN, >>> but I don't mind if this can be multi-valued for the multi-RDN >>> case. It is equal to the value of the RDN as presented in the RDN. >>> >>> This is not simply the subtype 'CN -> name', but a new attribute >>> unrelated to the existing definition of 'name'. >>> >>> I don't care what name is assigned to 'name', as I can easily remap >>> attributes. >>> >>> It would be great if this could be constructed such that it may be >>> declared to be unique for a particular one-level search (also an AD >>> requirement, but not one Samba4 requires or enforces at this time). >> The only problem I see in defining such an attribute is that its >> syntax should allow the value of any syntax, so it should probably be >> octetString or something like that. > > Why? IMO such an attribute type could be declared like any other > attribute type. If its syntax does not match the syntax of the > characteristic attribute invalidAttributeSyntax should be returned. Since we're dealing with something unspecified, you can: - be "strict": only allow the creation of this attribute when the naming attribute's syntax is 2.5.5.12 (unspecified, I understand, but we live in a wild world) - be "liberal": always allow the creation of this attribute by allowing it to contain any value > I wonder what the exact requirements for the implementation within slapd > are and some very rough ideas as food for thought: > > 1. slapd shall enforce uniqueness on one-level > => I'd vote for an additional feature of slapo-unique to define the > scope of uniqueness in a very flexible way > > 2. Value of 'name' has to match the value of the characteristic > attribute. Does this 1. has to be enforced within slapd or 2. could this > be enforced within smbd? > If 1. maybe the functionality/configuration of slapo-constraint could be > extended to define things like this. > (I've stumbled across ITS#5704. Isn't that already something like this?) not exactly, as sets do not allow a placeholder for the naming attribute; maybe something like "this/entryRDN.ava{0}.value" if it existed... p. > ...more...? > > In web2ldap's plugin classes I have implemented special treatment for > non-compliant notation for the DN part of LDAP URLs to reference a DN > based on an entry's DN: > > . the entry's DN > .. the entry's parent DN > _ the best matching namingContext for the entry's DN > > This strings can be appended to DNs. For example: > > 'ou=Users,_' would always refer to entry ou=Users below an arbitrary > naming context. > > 'ou=My Stuff,.' could refer to a container below a user's entry > > Maybe something like this could be helpful for an extended configuration > of slapo-unique and slapo-constraint. > > Hmm, also > > As said only very rough ideas... > > Ciao, Michael. > > Ing. Pierangelo Masarati OpenLDAP Core Team SysNet s.r.l. via Dossi, 8 - 27100 Pavia - ITALIA http://www.sys-net.it ----------------------------------- Office: +39 02 23998309 Mobile: +39 333 4963172 Fax: +39 0382 476497 Email: ando(a)sys-net.it -----------------------------------

1 0

Re: (ITS#6055) Samba4 need 'name' implementation like AD (RDN-Name)
by michael＠stroeder.com 18 Apr '09

18 Apr '09

ando(a)sys-net.it wrote: > ----- abartlet(a)samba.org wrote: >> Active Directory always presents an attribute 'name' that is always >> equal to the relative distinguished name. AD allows only one RDN, >> but I don't mind if this can be multi-valued for the multi-RDN >> case. It is equal to the value of the RDN as presented in the RDN. >> >> This is not simply the subtype 'CN -> name', but a new attribute >> unrelated to the existing definition of 'name'. >> >> I don't care what name is assigned to 'name', as I can easily remap >> attributes. >> >> It would be great if this could be constructed such that it may be >> declared to be unique for a particular one-level search (also an AD >> requirement, but not one Samba4 requires or enforces at this time). > > The only problem I see in defining such an attribute is that its > syntax should allow the value of any syntax, so it should probably be > octetString or something like that. Why? IMO such an attribute type could be declared like any other attribute type. If its syntax does not match the syntax of the characteristic attribute invalidAttributeSyntax should be returned. I wonder what the exact requirements for the implementation within slapd are and some very rough ideas as food for thought: 1. slapd shall enforce uniqueness on one-level => I'd vote for an additional feature of slapo-unique to define the scope of uniqueness in a very flexible way 2. Value of 'name' has to match the value of the characteristic attribute. Does this 1. has to be enforced within slapd or 2. could this be enforced within smbd? If 1. maybe the functionality/configuration of slapo-constraint could be extended to define things like this. (I've stumbled across ITS#5704. Isn't that already something like this?) ...more...? In web2ldap's plugin classes I have implemented special treatment for non-compliant notation for the DN part of LDAP URLs to reference a DN based on an entry's DN: . the entry's DN .. the entry's parent DN _ the best matching namingContext for the entry's DN This strings can be appended to DNs. For example: 'ou=Users,_' would always refer to entry ou=Users below an arbitrary naming context. 'ou=My Stuff,.' could refer to a container below a user's entry Maybe something like this could be helpful for an extended configuration of slapo-unique and slapo-constraint. Hmm, also As said only very rough ideas... Ciao, Michael.

1 0

Re: ITS#5860 slapd in openldap-2.4.16 still doesn't respect memory bounds
by sscdvp＠gmail.com 17 Apr '09

17 Apr '09

2009/4/17, sscdvp(a)gmail.com <sscdvp(a)gmail.com>: > 2009/4/17, sscdvp(a)gmail.com <sscdvp(a)gmail.com>: > > 2009/4/17, Quanah Gibson-Mount <quanah(a)zimbra.com>: > > > --On Friday, April 17, 2009 10:25 PM +0300 sscdvp(a)gmail.com wrote: > > > > Why do you have so many threads configured? How many cpu's do you have?? > > > > > > > > > > > > > > > > > > > > > I have tried openldap-2.4.16 (latest) BDB 4.7.25 with all patches from > > > > > Oracle on Dell AMD64 RAM 4Gb, 8 CPUs, OS Solaris 10. > > > > > > > > > > > > > I have an application which generates heavy load in slapd (many > > > > ldap_search'es and very very many ldap_add, ldap_modify, ldap_delete). > > > > BDB databases are situated on separate disk and BDB logs are on the other > > > > disk. But slapd performance (I mean time to answer) still cuts > > > > down during large disk IO periods 'cause these disks are used by other > > > > applications at the same time. And so I have increased number of threads > > > > from the defaults threads=32 up to 64, just to save some time for my > > > > application which requires not more then 30 ms for one search to > > > > complete. > > > > Just to be sure I have tried the default settings for threads and it's > > > > behaviour didn't changed - slapd heap memory just continuously grows (I > > > > have seen through pmap)... > > > > Also I have tried libtcmalloc which is usually recommended and solaris > > > > built in' libmtmalloc, but it didn't helped. > > > > > > > > > > Please copy replies to the bug tracker if you expect help. > > > > > > I will note that every thread consumes a large chunk of memory. I will also > > > note that the default is 16 threads, not 32. I will also note, that as far > > > as read operations are concerned, usually *fewer* threads is faster, not > > > more. > > >You appear to have quite an excessive amount of threads configured. > > > > > > I have mentioned that using the defaults didn't helped. Probably I was > mistaken about the default number of threads - but if keywords > "threads" is missing in config it simply means the defaults, right? > May be number of threads should be lower then the defaults (e.g. 6 or 8)? > Now I have 8 threads and slapd grows upto 2GB limit in 6 hours.

1 0

Re: ITS#5860 slapd in openldap-2.4.16 still doesn't respect memory bounds
by sscdvp＠gmail.com 17 Apr '09

17 Apr '09

2009/4/17, Quanah Gibson-Mount <quanah(a)zimbra.com>: > --On Friday, April 17, 2009 10:25 PM +0300 sscdvp(a)gmail.com wrote: > > Why do you have so many threads configured? How many cpu's do you have?? > > > > > > > > > > > I have tried openldap-2.4.16 (latest) BDB 4.7.25 with all patches from > > > Oracle on Dell AMD64 RAM 4Gb, 8 CPUs, OS Solaris 10. > > > > > > > I have an application which generates heavy load in slapd (many > > ldap_search'es and very very many ldap_add, ldap_modify, ldap_delete). > > BDB databases are situated on separate disk and BDB logs are on the other > > disk. But slapd performance (I mean time to answer) still cuts > > down during large disk IO periods 'cause these disks are used by other > > applications at the same time. And so I have increased number of threads > > from the defaults threads=32 up to 64, just to save some time for my > > application which requires not more then 30 ms for one search to > > complete. > > Just to be sure I have tried the default settings for threads and it's > > behaviour didn't changed - slapd heap memory just continuously grows (I > > have seen through pmap)... > > Also I have tried libtcmalloc which is usually recommended and solaris > > built in' libmtmalloc, but it didn't helped. > > > > Please copy replies to the bug tracker if you expect help. > > I will note that every thread consumes a large chunk of memory. I will also > note that the default is 16 threads, not 32. I will also note, that as far > as read operations are concerned, usually *fewer* threads is faster, not > more. >You appear to have quite an excessive amount of threads configured. > I have mentioned that using the defaults didn't helped. Probably I was mistaken about the default number of threads - but if keywords "threads" is missing in config it simply means the defaults, right? May be threads should be lower then the defaults (e.g. 6 or 8)?

1 0

Re: ITS#5860 slapd in openldap-2.4.16 still doesn't respect memory bounds
by Quanah Gibson-Mount 17 Apr '09

17 Apr '09

--On Friday, April 17, 2009 10:25 PM +0300 sscdvp(a)gmail.com wrote: > Why do you have so many threads configured? How many cpu's do you have?? > > > >> I have tried openldap-2.4.16 (latest) BDB 4.7.25 with all patches from >> Oracle on Dell AMD64 RAM 4Gb, 8 CPUs, OS Solaris 10. > > I have an application which generates heavy load in slapd (many > ldap_search'es and very very many ldap_add, ldap_modify, ldap_delete). > BDB databases are situated on separate disk and BDB logs are on the other > disk. But slapd performance (I mean time to answer) still cuts > down during large disk IO periods 'cause these disks are used by other > applications at the same time. And so I have increased number of threads > from the defaults threads=32 up to 64, just to save some time for my > application which requires not more then 30 ms for one search to > complete. > Just to be sure I have tried the default settings for threads and it's > behaviour didn't changed - slapd heap memory just continuously grows (I > have seen through pmap)... > Also I have tried libtcmalloc which is usually recommended and solaris > built in' libmtmalloc, but it didn't helped. Please copy replies to the bug tracker if you expect help. I will note that every thread consumes a large chunk of memory. I will also note that the default is 16 threads, not 32. I will also note, that as far as read operations are concerned, usually *fewer* threads is faster, not more. You appear to have quite an excessive amount of threads configured. Others will have to continue the conversation on what you're seeing growth wise. ;) --Quanah -- Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration

1 0

Re: ITS#5860 slapd in openldap-2.4.16 still doesn't respect memory bounds
by sscdvp＠gmail.com 17 Apr '09

17 Apr '09

2009/4/17, sscdvp(a)gmail.com <sscdvp(a)gmail.com>: > > > > > > > > > > I have tried openldap-2.4.16 (latest) BDB 4.7.25 with all patches from > > > Oracle on Dell AMD64 RAM 4Gb, 8 CPUs, OS Solaris 10. > > > According to ITS#5860 slapd should not allocate more memory than it was > > > specified. In fact I have the same behaviour as it was with previous > > > > > > version (2.4.11). It correctly allocates 1.2 Gb via shared memory (for > > > set_cachesize 300 MB). But slapd process continuosly grows in memory until > > > 32-bit limit then crashes due to unavalable memory resources. I need to > > > > > > mention bdb database size is less then 1 GB. > > > After restart it rapidly consumes 1.5 GB RSS in few minutes and then it > > > grows beyond the limit during 24 hours. The more queries are processing > > > the more memory it eats. > > > > > > I am using this database for logging which size grows. > > > The question is: is it possible to use 32-bit openldap with relatively > > > large database sizes that override 32-bit per-process memory space? > > > > > > My slapd.conf contains: > > > > > > > > > threads 64 > > > > > > > Why do you have so many threads configured? How many cpu's do you have?? > > > > I have tried openldap-2.4.16 (latest) BDB 4.7.25 with all patches from > > Oracle on Dell AMD64 RAM 4Gb, 8 CPUs, OS Solaris 10. > > I have an application which generates heavy load in slapd (many ldap_search'es and very very many ldap_add, ldap_modify, ldap_delete). BDB databases are situated on separate disk and BDB logs are on the other disk. But slapd performance (I mean time to answer) still cuts down during large disk IO periods 'cause these disks are used by other applications at the same time. And so I have increased number of threads from the defaults threads=32 up to 64, just to save some time for my application which requires not more then 30 ms for one search to complete. > Just to be sure I have tried the default settings for threads and it's behaviour didn't changed - slapd heap memory just continuously grows (I have seen through pmap)... > Also I have tried libtcmalloc which is usually recommended and solaris built in' libmtmalloc, but it didn't helped.

1 0

Re: (ITS#6058) slapcat aborts on a double free
by h.b.furuseth＠usit.uio.no 17 Apr '09

17 Apr '09

luca(a)openldap.org writes: > Slapcat on a slapd instance using rwm overlay, aborts with: > *** glibc detected *** free(): invalid pointer: 0x09ed06e3 *** It's an invalid free() of (static EntryHeader eh).e_bv.bv_val at: ch_free (ch_malloc.c:139) hdb_tool_entry_close (back-hdb/tools.c:164) slapcat (slapcat.c:135) main (main.c:636) already freed (as (Entry *e)->e_bv.bv_val) at: ch_free (ch_malloc.c:139) entry_clean (entry.c:498) entry_free (entry.c:514) rwm_entry_release_rw (overlays/rwm.c:827) overlay_entry_release_ov (backover.c:421) over_entry_release_rw (backover.c:463) be_entry_release_rw (backend.c:867) slapcat (slapcat.c:117) main (main.c:636) I needed to prepend include servers/slapd/schema/core.schema to your config. -- Hallvard

1 0

Re: (ITS#5975) minor issue with slapd.conf manpage
by h.b.furuseth＠usit.uio.no 17 Apr '09

17 Apr '09

eric(a)huss.org wrote: > There appears to be a couple slightly incorrect statements in the slapd.conf > manpage. It says that the argsfile and pidfile values are used "if started > without the debugging command line option". However, it doesn't appear to > make any difference whether or not you use the -d option with slapd. Fixed the doc in HEAD. Also the argsfile did not mention that the program name is written to argsfile as well as the options. -- Hallvard

1 0

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs