openldap-bugs

openldap-bugs@openldap.org

1 participants
20961 discussions

(ITS#4940) libldap doesn't wait for server's TLS close_notify
by guenther+ldapdev＠sendmail.com 30 Apr '07

30 Apr '07

Full_Name: Philip Guenther Version: 2.3.27 OS: Linux and Solaris URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (64.58.1.252) [I vaguely recall seeing a report of this issue in the archives of one of the mailing lists, but I can no longer find the original.] If you trace the packets sent when you use, for example, ldapsearch against a server on a different host, using either the -Z option to do TLS or using an ldaps URI, you'll discover that the TCP connection is actually reset instead of being closed cleanly: the client sends TCP RSTs in response to the server's final packets. This is because libldap uses the following sequence when unbind a TLS or SSL connection: 1) send the unbind request (over the TLS or SSL layer) 2) call SSL_shutdown(), sending the TLS close_notify alert 3) call close() After receiving the close_notify alert from step (2), the server sends back its own close_notify alert and then calls close(). However, because the client didn't wait for the server's response before calling close() on its end, the client's TCP stack considers the TCP connection to already be gone and responds with the RST packets. This occurs with Linux and Solaris clients and probably most other unices: the response to packets after a close() doesn't vary in my experience. There are a number of ways this can be handled: 1) change the client to wait until it sees the server's close_notify alert by replacing "SSL_shutdown( p->ssl );" in tls.c with the two lines: if (SSL_shutdown( p->ssl ) == 0) SSL_shutdown( p->ssl ); (I have confirmed that this works. As documented, the first call will return 1 if the server's close_notify has already been received, if not, the second call will block until it is received.) 2) change the client to not bother to send a close_notify alert when it's just going to close() the connection; change the server to not send a close_notify if it didn't get one. This probably violates the TLS spec, but the fact that TLS/1.1 permits resumption of sessions without close_notify having been sent indicates that the violation is not a major issue, particularly given that LDAP's unbind request prevents truncation attacks. Close_notifies are, of course, required if the client just wants to terminate the TLS layer and resume unprotected LDAP operations. 3) ignore the issue: it only causes one or two extra packets to be sent. While it also eliminates the TIME_WAIT state, LDAP's application-level close (the unbind request) means it doesn't need reliable full-duplex closure, so the only concern would be random connection issues from reincarnations of the TCP tuple, which is unlikely for an LDAP connection. Personally, I like the simplicity and cleanliness of solution (1).

1 0

(ITS#4939) Typographical error in client tool_common_usage()
by ceri＠submonkey.net 28 Apr '07

28 Apr '07

Full_Name: Ceri Davies Version: HEAD OS: FreeBSD URL: http://shrike.submonkey.net/~ceri/clients-tools-common.c.patch Submission from: (NULL) (81.106.128.65) The help output from the client tools contains a misspelling of "Identifier". See the patch at http://shrike.submonkey.net/~ceri/clients-tools-common.c.patch which applies against today's -HEAD. I seriously doubt that you need this, but: I, Ceri Davies, hereby place the following modifications to OpenLDAP Software (and only these modifications) into the public domain. Hence, these modifications may be freely used and/or redistributed for any purpose with or without attribution and/or other notice.

1 0

Re: (ITS#4938) hdb_db_close SEGVs
by quanah＠zimbra.com 26 Apr '07

26 Apr '07

----- quanah(a)zimbra.com wrote: > ----- richton(a)nbcs.rutgers.edu wrote: > > I don't have #5 (sleepycat#14657) nor the unofficial > > > http://www.stanford.edu/services/directory/openldap/configuration/patches/d… > > > > patch. As for the official one, I'm not sure about its relevance to > > the > > actual SEGV due to the "recovery...fail" comment. In other words, > > though > > it may be impacting the ability of alock/db_recover to do its thing, > > > that's just a side effect of the unclean shutdown which is the real > > bug > > here to my view. > > > > > Patch #5 specifically deals with a race condition where a checkpoint > is occuring while a cache buffer retrieval is also occuring causing a > database corruption that will later not be recoverable from. At > least, that's how I read sleepcat's description: > > 5. Fix a bug where cache buffer retrieval could race with a checkpoint > call, potentially causing database environment recovery to fail. > [#14657] > > Given that OpenLDAP checkpoints on shutdown, shutting down the server > could be what is triggering the issue for you. I'd suggest applying > the patch and seeing if this resolves your problem. Just to note, I shut down one of my 2.3.35 servers that's served out over 1 million connections since I brought it up, and everything was clean on both shutdown and startup. --Quanah

1 0

Re: (ITS#4938) hdb_db_close SEGVs
by quanah＠zimbra.com 26 Apr '07

26 Apr '07

----- richton(a)nbcs.rutgers.edu wrote: > I don't have #5 (sleepycat#14657) nor the unofficial > http://www.stanford.edu/services/directory/openldap/configuration/patches/d… > > patch. As for the official one, I'm not sure about its relevance to > the > actual SEGV due to the "recovery...fail" comment. In other words, > though > it may be impacting the ability of alock/db_recover to do its thing, > that's just a side effect of the unclean shutdown which is the real > bug > here to my view. Patch #5 specifically deals with a race condition where a checkpoint is occuring while a cache buffer retrieval is also occuring causing a database corruption that will later not be recoverable from. At least, that's how I read sleepcat's description: 5. Fix a bug where cache buffer retrieval could race with a checkpoint call, potentially causing database environment recovery to fail. [#14657] Given that OpenLDAP checkpoints on shutdown, shutting down the server could be what is triggering the issue for you. I'd suggest applying the patch and seeing if this resolves your problem. > The region size patch is interesting, but I will tell you that the > database in question has > > set_cachesize 0 200000000 0 > > and it (to a glance) looks like that only impacts the gig column, > which I > have as zero anyway. Yeah, the patch may not apply for you (I have a 3.5GB cache, so it does for me). Wouldn't harm anything, of course, if you decided later you needed a larger BDB cache. ;) --Quanah -- Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration

1 0

Re: (ITS#4938) hdb_db_close SEGVs
by richton＠nbcs.rutgers.edu 26 Apr '07

26 Apr '07

I don't have #5 (sleepycat#14657) nor the unofficial http://www.stanford.edu/services/directory/openldap/configuration/patches/d… patch. As for the official one, I'm not sure about its relevance to the actual SEGV due to the "recovery...fail" comment. In other words, though it may be impacting the ability of alock/db_recover to do its thing, that's just a side effect of the unclean shutdown which is the real bug here to my view. The region size patch is interesting, but I will tell you that the database in question has set_cachesize 0 200000000 0 and it (to a glance) looks like that only impacts the gig column, which I have as zero anyway. I can tell you that stop/starts weren't an issue with 2.3.32 and the same Sleepycat binaries...not that I stop/start often as a rule of thumb. (I am lately; we're implementing ando's {RADIUS} module.) But two identical traces on two different boxes caught my eye. On Thu, 26 Apr 2007, Quanah Gibson-Mount wrote: > > ----- richton(a)nbcs.rutgers.edu wrote: >> Full_Name: Aaron RIchton >> Version: 2.3.35 >> OS: Solaris 9 >> URL: ftp://ftp.openldap.org/incoming/ >> Submission from: (NULL) (128.6.30.206) >> >> >> BDB 4.2.52. I've had a couple (different) machines SEGV on slapd >> shutdown. Both >> had identical stack traces: > > Wierd, I've been running hdb on my servers for nearly a year without such an issue. Did this just start with 2.3.35? > > Also, what patches do you have applied to BDB 4.2.52. I'm up to 6 now. > > <http://www.stanford.edu/services/directory/openldap/configuration/bdb-build…> > > 5 are direct from sleepycat: > > <http://www.oracle.com/technology/products/berkeley-db/db/update/4.2.52/patc…> > > with the last one there possibly impacting you if you don't have it? > > --Quanah > > -- > Quanah Gibson-Mount > Principal Software Engineer > Zimbra, Inc > -------------------- > Zimbra :: the leader in open source messaging and collaboration > >

1 0

Re: (ITS#4938) hdb_db_close SEGVs
by quanah＠zimbra.com 26 Apr '07

26 Apr '07

----- richton(a)nbcs.rutgers.edu wrote: > Full_Name: Aaron RIchton > Version: 2.3.35 > OS: Solaris 9 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (128.6.30.206) > > > BDB 4.2.52. I've had a couple (different) machines SEGV on slapd > shutdown. Both > had identical stack traces: Wierd, I've been running hdb on my servers for nearly a year without such an issue. Did this just start with 2.3.35? Also, what patches do you have applied to BDB 4.2.52. I'm up to 6 now. <http://www.stanford.edu/services/directory/openldap/configuration/bdb-build…> 5 are direct from sleepycat: <http://www.oracle.com/technology/products/berkeley-db/db/update/4.2.52/patc…> with the last one there possibly impacting you if you don't have it? --Quanah -- Quanah Gibson-Mount Principal Software Engineer Zimbra, Inc -------------------- Zimbra :: the leader in open source messaging and collaboration

1 0

(ITS#4938) hdb_db_close SEGVs
by richton＠nbcs.rutgers.edu 26 Apr '07

26 Apr '07

Full_Name: Aaron RIchton Version: 2.3.35 OS: Solaris 9 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (128.6.30.206) BDB 4.2.52. I've had a couple (different) machines SEGV on slapd shutdown. Both had identical stack traces: current thread: t@1 =>[1] __dbreg_revoke_id(dbp = 0xa498818, have_lock = 0, force_id = -1), line 427 in "dbreg.c" [2] __dbreg_close_files(dbenv = 0x405320), line 206 in "dbreg_util.c" [3] __log_dbenv_refresh(dbenv = 0x405320), line 744 in "log.c" [4] __dbenv_refresh(dbenv = 0x405320, orig_flags = 0, rep_check = 0), line 648 in "env_open.c" [5] __dbenv_close(dbenv = 0x405320, rep_check = 0), line 579 in "env_open.c" [6] __dbenv_close_pp(dbenv = 0x405320, flags = 0), line 534 in "env_open.c" [7] hdb_db_close(be = 0x367340), line 517 in "init.c" [8] backend_shutdown(be = 0x367340), line 351 in "backend.c" [9] slap_shutdown(be = (nil)), line 279 in "init.c" [10] main(argc = 4, argv = 0xffbffd6c), line 870 in "main.c" No idea if this is Sleepycat or slapd, but it dirties the database in a way that automatic nor command-line db_recover appreciate: Ignoring log file: log.0000000007: magic number 0, not 40988 Invalid log file: log.0000000007: Invalid argument PANIC: Invalid argument PANIC: DB_RUNRECOVERY: Fatal error, run database recovery hdb_db_open: Database cannot be recovered, err -30978. Restore from backup! DB_ENV->lock_id_free interface requires an environment configured for the locking subsystem txn_checkpoint interface requires an environment configured for the transaction subsystem bdb_db_close: txn_checkpoint failed: Invalid argument (22) backend_startup_one: bi_db_open failed! (-30978)

1 0

Re: (ITS#4937) chain overlay bugs with Statslog & ldbm/ldif databases
by hyc＠symas.com 25 Apr '07

25 Apr '07

h.b.furuseth(a)usit.uio.no wrote: > Full_Name: Hallvard B Furuseth > Version: RE23 > OS: Linux > URL: > Submission from: (NULL) (129.240.202.105) > Submitted by: hallvard > > > overlay chain breaks with databases ldbm and ldif: > > ./run -b {ldbm or ldif} test032-chain outputs: > ... > Starting second slapd on TCP/IP port 9012... > ... > Comparing "cn=Mark Elliot,ou=Alumni Association,ou=People,dc=example,dc=com" > on port 9012... > ldapcompare failed (10)! > > ldapcompare output in testrun/test.out: > Compare Result: Referral (10) > Matched DN: ou=People,dc=example,dc=com > Referral: ldap://localhost:9011/ou=People,dc=example,dc=com > UNDEFINED > > testrun/slapd.1.log says the Compare was sent to the chained server, but > with baseDN=<Matched DN above>: > ... > do_compare: dn (ou=People,dc=example,dc=com) attr (cn) value (mark elliot) > conn=5 op=2 CMP dn="ou=People,dc=example,dc=com" attr="cn" > ... > conn=5 op=2 RESULT tag=111 err=16 text= > ... > > Slapcat says the Matched DN entry is correct (same as in a BDB run): > dn: ou=People,dc=example,dc=com > objectClass: referral > objectClass: extensibleObject > ou: People > ref: ldap://localhost:9011/ou=People,dc=example,dc=com > structuralObjectClass: referral > > > Also the Statslog output is wrong for all three databases: > > With LDBM and LDIF, the CMP operation is not logged, only its result: > grep '^conn=4' testrun/slapd.2.log # the ldapcompare operation above > conn=4 fd=14 ACCEPT from IP=127.0.0.1:45280 (IP=127.0.0.1:9012) > conn=4 op=0 BIND dn="" method=128 > conn=4 op=0 RESULT tag=97 err=0 text= > conn=4 op=1 RESULT tag=111 err=10 text= > conn=4 op=2 UNBIND > conn=4 fd=14 closed > > With BDB, the CMP operation is not logged, but both the returned > result and the suppressed referral result is logged: > conn=4 fd=12 ACCEPT from IP=127.0.0.1:42190 (IP=127.0.0.1:9012) > conn=4 op=0 BIND dn="" method=128 > conn=4 op=0 RESULT tag=97 err=0 text= > conn=4 op=1 RESULT tag=111 err=6 text= > conn=4 op=1 RESULT tag=111 err=10 text= > conn=4 op=2 UNBIND > conn=4 fd=12 closed > > > I haven't tried other databases, but since two different databases > had the same bug and same wrong Statlog, I'm guessing others do too. This seems to be a known limitation, see the FIXME in slapd/compare.c, near SLAP_COMPARE_IN_FRONTEND. At least, that explains the problem for back-ldif, which does not provide its own compare entry point. As for back-ldbm, there is a missing referral_rewrite call in back-ldbm/referral.c. I don't consider it worthwhile to fix back-ldbm, but you're welcome to fix it if you wish. Just have a look at back-bdb/referral.c for the correct code. -- -- 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#4937) chain overlay bugs with Statslog & ldbm/ldif databases
by h.b.furuseth＠usit.uio.no 25 Apr '07

25 Apr '07

Full_Name: Hallvard B Furuseth Version: RE23 OS: Linux URL: Submission from: (NULL) (129.240.202.105) Submitted by: hallvard overlay chain breaks with databases ldbm and ldif: ./run -b {ldbm or ldif} test032-chain outputs: ... Starting second slapd on TCP/IP port 9012... ... Comparing "cn=Mark Elliot,ou=Alumni Association,ou=People,dc=example,dc=com" on port 9012... ldapcompare failed (10)! ldapcompare output in testrun/test.out: Compare Result: Referral (10) Matched DN: ou=People,dc=example,dc=com Referral: ldap://localhost:9011/ou=People,dc=example,dc=com UNDEFINED testrun/slapd.1.log says the Compare was sent to the chained server, but with baseDN=<Matched DN above>: ... do_compare: dn (ou=People,dc=example,dc=com) attr (cn) value (mark elliot) conn=5 op=2 CMP dn="ou=People,dc=example,dc=com" attr="cn" ... conn=5 op=2 RESULT tag=111 err=16 text= ... Slapcat says the Matched DN entry is correct (same as in a BDB run): dn: ou=People,dc=example,dc=com objectClass: referral objectClass: extensibleObject ou: People ref: ldap://localhost:9011/ou=People,dc=example,dc=com structuralObjectClass: referral Also the Statslog output is wrong for all three databases: With LDBM and LDIF, the CMP operation is not logged, only its result: grep '^conn=4' testrun/slapd.2.log # the ldapcompare operation above conn=4 fd=14 ACCEPT from IP=127.0.0.1:45280 (IP=127.0.0.1:9012) conn=4 op=0 BIND dn="" method=128 conn=4 op=0 RESULT tag=97 err=0 text= conn=4 op=1 RESULT tag=111 err=10 text= conn=4 op=2 UNBIND conn=4 fd=14 closed With BDB, the CMP operation is not logged, but both the returned result and the suppressed referral result is logged: conn=4 fd=12 ACCEPT from IP=127.0.0.1:42190 (IP=127.0.0.1:9012) conn=4 op=0 BIND dn="" method=128 conn=4 op=0 RESULT tag=97 err=0 text= conn=4 op=1 RESULT tag=111 err=6 text= conn=4 op=1 RESULT tag=111 err=10 text= conn=4 op=2 UNBIND conn=4 fd=12 closed I haven't tried other databases, but since two different databases had the same bug and same wrong Statlog, I'm guessing others do too.

1 0

Re: (ITS#4936) default modulepath
by ahasenack＠terra.com.br 23 Apr '07

23 Apr '07

On Mon, Apr 23, 2007 at 09:17:26PM +0000, abartlet(a)samba.org wrote: > It would be very useful if OpenLDAP would provide a default value for > modulepath. > > This would ensure that on systems with multiple openldap installs (such as test > rigs) that we do not need to specify the *correct* modulepath, when starting the > config with different slapd binaries. > > Similarly, it would avoid the need to guess which modulepath is correct for a > given system, such as when we find slapd in the path (as Samba4's test mode > does), but needs to automatically enable modules. Another scenario is with some tool providing a config file for slapd. It wouldn't have to guess the host arch whether it's i386 or x86_64 and set the correct module_path (/usr/lib/openldap vs /usr/lib64/openldap).

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