openldap-bugs

openldap-bugs@openldap.org

1 participants
20960 discussions

Re: (ITS#5016) slapd generates unjustified error 3 (Time Limit Exceeded) at checkpoint time
by ando＠sys-net.it 14 Jun '07

14 Jun '07

ali.pouya(a)free.fr wrote: > What bothers me is that the regular client requests to the consumer receive > (scarcely and randomly) "Time Limit Exceeded" at checkpoint time. Would it be possible to get, from the consumer's log, what was the actually requested time limit in those cases? This means: track down what connection/operation was concluded by a "Time Limit Exceeded" response; then find out the corresponding request, and see what was the requested time limit. You need "args" debug level for this purpose, and you need to look at the second of the three numbers that end the SRCH line in the logs; for example: SRCH "dc=example,dc=com" 2 0 20 5 0 filter: (uid=foobar) attrs: 1.1 means: search "dc=example,dc=com" with scope subtree (2) and never deref aliases (0), using a size limit of 20, a timelimit of 5 and no attrsonly (0). p. 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 Email: pierangelo.masarati(a)sys-net.it ---------------------------------------

1 0

Re: (ITS#4943) tpool.c pause vs. finish
by h.b.furuseth＠usit.uio.no 14 Jun '07

14 Jun '07

Howard Chu writes: > Hallvard B Furuseth wrote: >> hyc(a)symas.com writes: >>> h.b.furuseth(a)usit.uio.no wrote: >>>> * When a thread is finishing, make it go active for context_reset(). >>>> Should make slapi module destructors safer, and I don't see that it >>>> can hurt. >>> I don't understand this. >> >> Let me dig a bit back in the ITS... >> "context_reset() can call connection_fake_destroy(), which via >> slapi_int_free_object_extensions() calls slapi module destructors, >> which can do who-knows-what." > > context_reset() can only be called from the main thread. It can (must) > never be called by anything running from the pool. It's called from pool_wrapper(). pool_wrapper() could go active so there can be no pause while that context_reset() call is running. >>>> * Scheduling: If several threads call pool_pause(), (...) (need to check out how it should work...) >>>> - pool_context() breaks if several ldap_pvt_thread_t bit patterns can >>>> represent the same thread ID: TID_HASH() would give different hashes >>>> for the same thread, and pool_context() stops searching when it hits >>>> a slot with ctx == NULL. (My previous bug report was a bit wrong.) >>> >>> How can a single thread ID be represented by multiple bit patterns? >> >> A struct/union with padding bytes seems the least unlikely possibility. Note that I did say "least unlikely" rather than "most likely": It's not that I expect to encounter it, more that I'd prefer not to be surprised in the form of a slapd malfuntion. So in that respect, my preference (not for RE23) is to cast to an integer for the hash, and also to add pthread_<set/get>specific() so thread_keys[] and its hash will only be a fallback solution. > Any implementation that leaves uninitialized padding bytes would be a > bug. Tools like valgrind would barf all over them. No. Valid compiler, valid code. Therefore valgrind is silent about copying uninitialized values, it only complains when you use them for something - like write() or arithmetic. So a valgrind complaint about an uninitialized _value_ isn't necessary at the point the uninitialized _variable_ is read. E.g. this complains in printf(), not in foo(): int foo() { struct { char c; /*maybe padding here*/ int j; } s; s.c = 1; s.j = 2; return ((unsigned char *)&s)[1]; /* maybe padding byte */ } int main() { printf("%d\n", foo()); } >> A pointer in hardware where several address representations map to the >> same physical address, and the compiler handles that by normalizing >> pointers when they are compared/subtracted. Like DOS "huge" pointers >> would have been if the compiler normalized them when comparing them >> instead of when incrementing/decrementing them. 20-bit address bus, >> 32-bit pointers, physical address = 16 * <segment half of pointer> + >> <offset half of pointer>. > > Nobody voluntarily uses such memory models today. The very mention of "DOS" > invalidates this discussion since there is no threading environment there. Right, at this point I answered how it _can_ happen, not why I expect to encounter it. I could see the demands of hardware design cause strange memory models to proliferate again, but I'm not holding my breath. I'd be a bit less surprised to encounter integers with padding bits again, but not for a simple ID type, and hashing an integer instead of its bytes would fix it anyway. >>> The best fix may be to just use the address of the thread's stack as the >>> thread ID. That's actually perfect for our purpose. >> >> How does a function called from somewhere inside the thread know that >> address? That's almost what the user context is, and thus what >> thread_keys[] maps the thread ID to. > > It requires some degree of machine-dependence, which is why I never > bothered to code it. The principle is simple though - all of our thread > stacks are some multiple of 1MB in size. Take the address of any local > variable, mask off the low 20 bits, and you have a unique thread ID (+/- > some additional masking/shifting based on the actual thread stack size). I'm not sure how to determine _which_ multiple of 1MB - start two threads and compare their stack addresses, maybe? In any case, I agree it sounds rather more machine-dependent than the current code. -- Regards, Hallvard

1 0

Re: (ITS#5016) slapd generates unjustified error 3 (Time Limit Exceeded) at checkpoint time
by ali.pouya＠free.fr 14 Jun '07

14 Jun '07

Selon Pierangelo Masarati <ando(a)sys-net.it>: > ali.pouya(a)free.fr wrote: > > > Hi, > > I use openldap 2.3.35 with the following bdb configuration. > > > > database bdb > > .... > > checkpoint 1000000 10 > > > > In my DB_CONFIG file I have : > > > > set_flags DB_LOG_AUTOREMOVE > > > > There ise a great LDAP read/write activity on my server. > > > > I notice that every 10 minutes some of my search requests receive randomly > an > > error 3 (Time Limit Exceeded) within one or two seconds while I'm using the > > slapd default timelimit (3600 seconds). > > > > I think a relation exists between this error and the checkpoint activity > (slapd > > is inclined to generate unjustified error 3 at checkpoint time) . > > Is it possible to study this problem ? > > > > I can send more information if required. > > > Do you mean that regular requests (to the producer or to the consumer?) > often receive "Time Limit Exceeded", or that the replication requests > issued by the consumer receive "Time Limit Exceeded" from the producer? > > p. > > > > 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 > Email: pierangelo.masarati(a)sys-net.it > --------------------------------------- > > > Hi Pierangelo; What bothers me is that the regular client requests to the consumer receive (scarcely and randomly) "Time Limit Exceeded" at checkpoint time. The replication requests to the master also receive this error after 3600 seconds when I'm synchronizing a new empty replica (as though I set timelimit unlimited for the sync client), but this is not a problem for me because the syncrepl client makes a new connection to the master and everything works well. Best regards Ali

1 0

Re: (ITS#5016) slapd generates unjustified error 3 (Time Limit Exceeded) at checkpoint time
by ando＠sys-net.it 14 Jun '07

14 Jun '07

ali.pouya(a)free.fr wrote: > Hi, > I use openldap 2.3.35 with the following bdb configuration. > > database bdb > .... > checkpoint 1000000 10 > > In my DB_CONFIG file I have : > > set_flags DB_LOG_AUTOREMOVE > > There ise a great LDAP read/write activity on my server. > > I notice that every 10 minutes some of my search requests receive randomly an > error 3 (Time Limit Exceeded) within one or two seconds while I'm using the > slapd default timelimit (3600 seconds). > > I think a relation exists between this error and the checkpoint activity (slapd > is inclined to generate unjustified error 3 at checkpoint time) . > Is it possible to study this problem ? > > I can send more information if required. Do you mean that regular requests (to the producer or to the consumer?) often receive "Time Limit Exceeded", or that the replication requests issued by the consumer receive "Time Limit Exceeded" from the producer? p. 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 Email: pierangelo.masarati(a)sys-net.it ---------------------------------------

1 0

(ITS#5018) memory leak
by niloulili＠hotmail.com 14 Jun '07

14 Jun '07

Full_Name: nina wang Version: 2.2.19 OS: Vxworks mips URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (222.68.187.35) I remove these code from linux to gcc. a computer equipmented os win 2000 server acts as server setting up AD . i start a thread to init a ldap connection and bind them with a valid account and password,configured in server. then unbind them every several seconds. through special tools ,i can see that free memory always being decreasing. if i only do action ldap_init,without bind to server,no the said scene . THE CODE IS : { LDAP *ld; LDAPMessage *res=NULL, *e; int i; char *a, *dn; BerElement *ptr; char **vals; /* open a connection */ if ( (ld = ldap_init("192.xxx.xxx.xxx",389 ))== NULL ) return 1; if ( ldap_simple_bind_s( ld,"CN=nina wang,CN=Users,DC=ldap,DC=qtier,DC=com","0000" ) != LDAP_SUCCESS ) { ldap_perror( ld, "ldap_simple_bind_s" ); return 1; } ldap_msgfree( res ); /* close and free connection resources */ ldap_unbind( ld ); return 0; } Hope to receive your response soon. Thank you .

1 0

Re: (ITS#5017) slapd generates unjustified error 3 (Time Limit Exceeded) at check
by ali.pouya＠free.fr 13 Jun '07

13 Jun '07

This ITS is a duplicate of ITS 5016. It is to be destroyed. Sorry for mis handling. Ali

1 0

(ITS#5017) slapd generates unjustified error 3 (Time Limit Exceeded) at checkpoint time
by ali.pouya＠free.fr 13 Jun '07

13 Jun '07

Full_Name: Ali Pouya Version: 2.3.35 OS: Linux RedHat AS4 (kernel 2.6) URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (145.242.3.30) Hi, I use openldap 2.3.35 with the following bdb configuration. database bdb .... checkpoint 1000000 10 In my DB_CONFIG file I have : set_flags DB_LOG_AUTOREMOVE There ise a great LDAP read/write activity on my server. I notice that every 10 minutes some of my search requests receive randomly an error 3 (Time Limit Exceeded) within one or two seconds while I'm using the slapd default timelimit (3600 seconds). I think a relation exists between this error and the checkpoint activity (slapd is inclined to generate unjustified error 3 at checkpoint time) . Is it possible to study this problem ? I can send more information if required. Thanks for your help Best regards Ali Pouya

1 0

(ITS#5016) slapd generates unjustified error 3 (Time Limit Exceeded) at checkpoint time
by ali.pouya＠free.fr 13 Jun '07

13 Jun '07

1 0

(ITS#5015) pcache caches results of aborted searches
by rhafer＠suse.de 13 Jun '07

13 Jun '07

Full_Name: Ralf Haferkamp Version: RE23 OS: URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (89.166.165.45) Pcache will store the results of a search that got aborted (e.g. by hitting Ctrl+c during ldapsearch, oder using ldap_abandon) in it database. The next time the some query is issued it will answer the query from the cache, resulting in a incomplete result set. HEAD doesn't show this problem. Applying the changes from result.c r.1.289 to RE23 seems to fix this problem.

1 0

RE: Native Win32 build with WGCC/Interix (ITS#4825)
by Markus.Duft＠salomon.at 13 Jun '07

13 Jun '07

Hi! First thanks for reviewing the patch! Howard Chu <mailto:openldap-its@OpenLDAP.org> wrote: > I have not tested this patch, but it does not appear to be correct. > > In ldap_cdefs.h, you should not be testing the PIC macro since that > controls whether the current source file is being built as > position-independent. The macro you actually need to test is > LDAP_LIBS_DYNAMIC, just as the regular WIN32 case tests. Note that > the current macros work fine for both MSVC and MinGW/gcc; I don't see > why you've chosen to change the LBER_F definitions for your wgcc. Uh.. I'd have to look into that again ;o) Still this works fine since some time now, and the libraries are under heavy usage at our company... > > Likewise, your hack to portable.hin should not be necessary, as all > of the places that use sockets or timevals already include the > appropriate header files. Again, i'd have to look at this again... > > The patch to ure.h seems unnecessary, since the file exists at both > locations (the original, and via symlink). Smylinks are not supported when windows building on interix ;o// This caused some troubles as far as i remember. > > Finally, unless you're using a flag to force all symbols to be > exported, I think omitting slapd.exp in the slapd/Makefile.in is a > mistake. WGCC automatically exports/imports correctly. > > Really, the comments in ldap_cdefs.h are pretty explicit about why > each piece of the current Windows support is as it is. Your patch > seems to preclude some of the combinations of DLL/static linking that > we went to great pains to support. WGCC claims to support (with this patch) *all* possible static/shared combinations (without pain ;o)). WGCC generates code to support static libraries, even if it was compiled like a shared library (with __declspec(dllimport)'s, etc.), so most of the painfull stuff is not needed. On the other hand to support *all* configurations, one needs to allway dllimport, so both shared and static work ... > > For the include/Makefile.in it would probably have made more sense to > define a WINPATH macro to either "cygpath -w" or "unixpath2win" and > use that, rather than repeating all of the SED invocations. Yes, it would have been better, but it's a long time since i wrote the patch, and now i'd propably do more things differently ;o) > > These days I wonder whether it wouldn't be smarter to just remove all > of the special casing for braindead backslash directory separators. > Since Windows now supports forward slashes everywhere, it would make > life easier to just use forward slashes consistently throughout. Would be cool, but still path handling functions from windows return paths with backslashes, so those would need to be handled too.. Cheers, Markus -- 20. Juni 2007 Salomon Automation am 2. Schweizerischen Supply Chain Management Forum der GS1 Schweiz in Baden. Tagungsort: Trafohalle Baden, Schweiz 20. Juni 2007 MoveRetail-Handelstag, Hamburg Die MoveRetail Partner, Salomon Automation, maxess systemhaus, Superdata, Remira, POS Systemhaus und Mosaic veranstalten den 1. MoveRetail-Handelstag. Tagungsort: Steigenberger Hotel, Hamburg 28. Juni 2007 6. BVL Logistiktag Steiermark in der WAMAS City, Friesach bei Graz "Trends in der Logistik ": hoch automatisierte Systeme, mobile Waren- und Leergutverfolgung im Lebensmittelhandel und Generalunternehmerschaft versus Einzelprojekte. Termin: 28. Juni 2007, ab 15.00 Uhr Ort: Salomon Automation, Friesachstra?e 15, 8114 Friesach bei Graz, Osterreich Salomon Automation GmbH - Friesachstrasse 15 - A-8114 Friesach bei Graz Sitz der Gesellschaft: Friesach bei Graz UID-NR:ATU28654300 - Firmenbuchnummer: 49324 K Firmenbuchgericht: Landesgericht fur Zivilrechtssachen Graz

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