openldap-bugs

openldap-bugs@openldap.org

1 participants
20961 discussions

(ITS#4919) Why CPU usage can not reach 100% while doing modify operation?
by seuler＠163.com 09 Apr '07

09 Apr '07

Full_Name: Seuler Shi Version: openldap-2.3.30 OS: SUSE 9 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (194.138.202.50) When I do the search operation, the CPU usage of OpenLDAP Server can reach 100% when the number of search operations per second arise? But for the modify operation, the CPU usage can only reach 87% at most? What is the bottleneck that makes OpenLDAP can not make CPU recource fully while doing modify operation?

1 0

Re: (ITS#4916) syncrpl consumer not listing contextcsn as operation attr
by brian＠zimbra.com 07 Apr '07

07 Apr '07

thanks, that worked, this case can be closed. Brian Peterson On Apr 6, 2007, at 1:58 PM, Howard Chu wrote: > brian(a)zimbra.com wrote: >> Full_Name: Brian Peterson >> Version: 2.3.34 >> OS: Linux >> URL: pro >> Submission from: (NULL) (66.92.25.194) >> When setup sync replication the contextCSN is available via >> ldapsearch on the >> provider but not the consumer as an operation attribute. If we do >> a slapcat on >> the consumer the resulting ldif contains the contextCSN. > > This is because your context entry has no DN, it's a glue entry. > When you do a search on it you get the contents of the server > rootDSE instead of your glue entry. The search works on the > provider because the syncprov overlay intercepts searches to insert > the current contextCSN value in the result. You can add a syncprov > overlay to the consumer to get the same effect. > >> Top entry from slapcat ldif on the consumer. >> dn: >> objectClass: glue >> structuralObjectClass: glue >> contextCSN: 20070405182202Z#000000#00#000000 >> entryCSN: 20070405182202Z#000000#00#000000 >> modifiersName: uid=zimbra,cn=admins,cn=zimbra >> modifyTimestamp: 20070405182202Z > > > -- > -- Howard Chu > Chief Architect, Symas Corp. http://www.symas.com > Director, Highland Sun http://highlandsun.com/hyc/ > Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

Re: (ITS#4918) int = pointer assignment in syncrepl
by hyc＠symas.com 06 Apr '07

06 Apr '07

h.b.furuseth(a)usit.uio.no wrote: > Full_Name: Hallvard B Furuseth > Version: RE23 > OS: > URL: > Submission from: (NULL) (129.240.202.105) > Submitted by: hallvard > > > In RE23:syncrepl.c:do_syncrepl() lin 1155, the statement > fail = si->si_retrynum; > assigns an int* to an int. > > > . > Thanks, fixed. -- -- Howard Chu Chief Architect, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

(ITS#4918) int = pointer assignment in syncrepl
by h.b.furuseth＠usit.uio.no 06 Apr '07

06 Apr '07

Full_Name: Hallvard B Furuseth Version: RE23 OS: URL: Submission from: (NULL) (129.240.202.105) Submitted by: hallvard In RE23:syncrepl.c:do_syncrepl() lin 1155, the statement fail = si->si_retrynum; assigns an int* to an int.

1 0

Re: (ITS#4916) syncrpl consumer not listing contextcsn as operation attr
by hyc＠symas.com 06 Apr '07

06 Apr '07

brian(a)zimbra.com wrote: > Full_Name: Brian Peterson > Version: 2.3.34 > OS: Linux > URL: pro > Submission from: (NULL) (66.92.25.194) > > > When setup sync replication the contextCSN is available via ldapsearch on the > provider but not the consumer as an operation attribute. If we do a slapcat on > the consumer the resulting ldif contains the contextCSN. This is because your context entry has no DN, it's a glue entry. When you do a search on it you get the contents of the server rootDSE instead of your glue entry. The search works on the provider because the syncprov overlay intercepts searches to insert the current contextCSN value in the result. You can add a syncprov overlay to the consumer to get the same effect. > Top entry from slapcat ldif on the consumer. > dn: > objectClass: glue > structuralObjectClass: glue > contextCSN: 20070405182202Z#000000#00#000000 > entryCSN: 20070405182202Z#000000#00#000000 > modifiersName: uid=zimbra,cn=admins,cn=zimbra > modifyTimestamp: 20070405182202Z > > > -- -- Howard Chu Chief Architect, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

Re: (ITS#4917) at_insert() moves sat_ad_mutex
by h.b.furuseth＠usit.uio.no 06 Apr '07

06 Apr '07

hyc(a)symas.com wrote: >h.b.furuseth(a)usit.uio.no wrote: >> If so, the simplest clean fix may be to destroy and reinitialize >> the mutexes (instead of swapping them back), I think strictly >> speaking behaviour is undefined if one copies mutexes around at >> all. Or since one will immediately be destroyed, maybe they should >> be destroyed when SLAP_AT_DELETED is set, if it is not needed then. > > Destroying immediately is not feasible since the attribute may still > be in use. OK. > Show me the POSIX spec that indicates this behavior is undefined. I wasn't thinking of POSIX in particular (OpenLDAP supports several mutex types afer all), though the pthread_mutex_init() spec says Only mutex itself may be used for performing synchronization. The result of referring to copies of mutex in calls to pthread_mutex_lock(), pthread_mutex_trylock(), pthread_mutex_unlock(), and pthread_mutex_destroy() is undefined. Attempting to initialize an already initialized mutex results in undefined behavior. The current swap definitely makes it undefined. Whether swapping and copying back does so as well is probably just pedantry when they are unlocked, though in theory I imagine some thread magic could happen even to "passive" mutexes. Except I take it only one thread is running... -- Regards, Hallvard

1 0

Re: (ITS#4917) at_insert() moves sat_ad_mutex
by hyc＠symas.com 06 Apr '07

06 Apr '07

h.b.furuseth(a)usit.uio.no wrote: > Full_Name: Hallvard B Furuseth > Version: HEAD > OS: > URL: > Submission from: (NULL) (129.240.202.105) > Submitted by: hallvard > > > slapd/at.c:at_insert() swaps the contents of the initialized > mutexes 'old_sat->sat_ad_mutex' and 'sat->sat_ad_mutex' in the > /* replacing a deleted definition? */ code. > > I hope they are at least unlocked at the time... > Does ldap_pvt_thread_pool_pause() in config_back_add() & co block > the other threads? Yes of course, otherwise none of the config activities would be safe. > If so, the simplest clean fix may be to destroy and reinitialize > the mutexes (instead of swapping them back), I think strictly > speaking behaviour is undefined if one copies mutexes around at > all. Or since one will immediately be destroyed, maybe they should > be destroyed when SLAP_AT_DELETED is set, if it is not needed then. Destroying immediately is not feasible since the attribute may still be in use. Show me the POSIX spec that indicates this behavior is undefined. -- -- Howard Chu Chief Architect, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

(ITS#4917) at_insert() moves sat_ad_mutex
by h.b.furuseth＠usit.uio.no 06 Apr '07

06 Apr '07

Full_Name: Hallvard B Furuseth Version: HEAD OS: URL: Submission from: (NULL) (129.240.202.105) Submitted by: hallvard slapd/at.c:at_insert() swaps the contents of the initialized mutexes 'old_sat->sat_ad_mutex' and 'sat->sat_ad_mutex' in the /* replacing a deleted definition? */ code. I hope they are at least unlocked at the time... Does ldap_pvt_thread_pool_pause() in config_back_add() & co block the other threads? If so, the simplest clean fix may be to destroy and reinitialize the mutexes (instead of swapping them back), I think strictly speaking behaviour is undefined if one copies mutexes around at all. Or since one will immediately be destroyed, maybe they should be destroyed when SLAP_AT_DELETED is set, if it is not needed then.

1 0

(ITS#4916) syncrpl consumer not listing contextcsn as operation attr
by brian＠zimbra.com 06 Apr '07

06 Apr '07

Full_Name: Brian Peterson Version: 2.3.34 OS: Linux URL: pro Submission from: (NULL) (66.92.25.194) When setup sync replication the contextCSN is available via ldapsearch on the provider but not the consumer as an operation attribute. If we do a slapcat on the consumer the resulting ldif contains the contextCSN. relavent slapd.conf entries from provider(please let me know if you want the entire file). database bdb suffix "" rootdn "uid=zimbra,cn=admins,cn=zimbra" # recommended for replication index entryUUID eq index entryCSN eq overlay syncprov syncprov-checkpoint 100 10 syncprov-sessionlog 500 slapd.conf entries from consumer. database bdb suffix "" rootdn "uid=zimbra,cn=admins,cn=zimbra" syncrepl rid=022 provider=ldap://ldapmaster:389 retry="60 +" type=refreshAndPersist schemachecking=off updatedn="uid=zimbrareplication,cn=admins,cn=zimbra" searchbase="" bindmethod=simple binddn="uid=zimbra,cn=admins,cn=zimbra" creditionals=secret updateref ldap://ldapmaster:389 Ldap search output from the consumer ldapsearch -LLL -x -D uid=zimbra,cn=admins,cn=zimbra -w secret -H ldap://ldapreplica:389 -s base -b "" + dn: structuralObjectClass: OpenLDAProotDSE configContext: cn=config namingContexts: supportedControl: 1.3.6.1.4.1.4203.1.9.1.1 supportedControl: 2.16.840.1.113730.3.4.18 supportedControl: 2.16.840.1.113730.3.4.2 supportedControl: 1.3.6.1.4.1.4203.1.10.1 supportedControl: 1.2.840.113556.1.4.319 supportedControl: 1.2.826.0.1.334810.2.3 supportedControl: 1.2.826.0.1.3344810.2.3 supportedControl: 1.3.6.1.1.13.2 supportedControl: 1.3.6.1.1.13.1 supportedControl: 1.3.6.1.1.12 supportedExtension: 1.3.6.1.4.1.1466.20037 supportedExtension: 1.3.6.1.4.1.4203.1.11.1 supportedExtension: 1.3.6.1.4.1.4203.1.11.3 supportedFeatures: 1.3.6.1.1.14 supportedFeatures: 1.3.6.1.4.1.4203.1.5.1 supportedFeatures: 1.3.6.1.4.1.4203.1.5.2 supportedFeatures: 1.3.6.1.4.1.4203.1.5.3 supportedFeatures: 1.3.6.1.4.1.4203.1.5.4 supportedFeatures: 1.3.6.1.4.1.4203.1.5.5 supportedLDAPVersion: 3 entryDN: subschemaSubentry: cn=Subschema Top entry from slapcat ldif on the consumer. dn: objectClass: glue structuralObjectClass: glue contextCSN: 20070405182202Z#000000#00#000000 entryCSN: 20070405182202Z#000000#00#000000 modifiersName: uid=zimbra,cn=admins,cn=zimbra modifyTimestamp: 20070405182202Z

1 0

Re: (ITS#4915) Syncrepl task exits are not logged
by hyc＠symas.com 06 Apr '07

06 Apr '07

unix.gurus(a)gmail.com wrote: > Full_Name: Sean Burford > Version: 2.3.32 > OS: Linux > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (65.57.245.11) > > > It is hard to tell when the syncrepl task has exited. The attached patch adds > logging of syncrepl hard and soft failures. > > Using this patch, it is easier to spot unfortunate conditions where recoverable > failures are treated as hard failures, for example the third line below shows > that a (temporarily) unavailable kerberos credential cache causes a hard > failure: > GSSAPI Error: Miscellaneous failure (No credentials cache found) > do_syncrep1: rid 123 ldap_sasl_interactive_bind_s failed (-2) > do_syncrepl: rid 123 hard failure Looks like a good idea, thanks. I think I'd just change the messages to "giving up" or "quitting", and "retrying". That final else clause in your patch should never occur, there is no 3rd outcome. > > Hmm, no attachment because the FTP server returns: > 200 PORT command successful. > 150 Opening ASCII mode data connection for > 'openldap-2.3.32-syncrepl-log-failure.patch'. > ## > 452 Error writing file: No space left on device. > 1104 bytes sent in 0.00011 seconds (1e+04 Kbytes/s) > > The patch was: > *** openldap-2.3.32.orig/servers/slapd/syncrepl.c Wed Apr 4 17:21:31 > 2007 > --- openldap-2.3.32/servers/slapd/syncrepl.c Wed Apr 4 17:55:58 2007 > *************** > *** 1147,1159 **** > --- 1147,1169 ---- > } > > if ( !si->si_retrynum || si->si_retrynum[i] == RETRYNUM_TAIL ) > { > + Debug( LDAP_DEBUG_ANY, > + "do_syncrepl: rid %03d hard failure\n", > + si->si_rid, 0, 0 ); > ldap_pvt_runqueue_remove( &slapd_rq, rtask ); > } else if ( RETRYNUM_VALID( si->si_retrynum[i] ) ) { > if ( si->si_retrynum[i] > 0 ) > si->si_retrynum[i]--; > + Debug( LDAP_DEBUG_ANY, > + "do_syncrepl: rid %03d soft failure, > rescheduling (retries left: %d)\n", > + si->si_rid, si->si_retrynum[i], 0 ); > rtask->interval.tv_sec = si->si_retryinterval[i]; > ldap_pvt_runqueue_resched( &slapd_rq, rtask, 0 ); > slap_wake_listener(); > + } else { > + Debug( LDAP_DEBUG_ANY, > + "do_syncrepl: rid %03d soft failure, > rescheduling\n", > + si->si_rid, 0, 0 ); > } > } > > > -- -- Howard Chu Chief Architect, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

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