openldap-bugs April 2012

openldap-bugs@openldap.org

36 participants
172 discussions

(ITS#7230) back-mdb: slapadd fails when not run in quickmode
by quanah＠OpenLDAP.org 05 Apr '12

05 Apr '12

Full_Name: Quanah Gibson-Mount Version: RE24 4/4/2012 OS: Linux 2.6 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (75.108.184.39) Running slapadd without quickmode quickly results in failure: zimbra@zre-ldap001:~$ time /opt/zimbra/openldap/sbin/slapadd -b "" -l /tmp/ldap.bak -F /opt/zimbra/data/ldap/config * 0.63% eta 01h08m elapsed 26s spd 1.1 M/s slapadd: ./../../../libraries/libmdb/mdb.c:3459: mdb_page_search: Assertion `root > 1' failed. Aborted

1 0

(ITS#7229) back-mdb: unable to load large databases
by quanah＠OpenLDAP.org 05 Apr '12

05 Apr '12

Full_Name: Quanah Gibson-Mount Version: RE24 4/4/2012 OS: Linux 2.6 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (75.108.184.39) When loading a large database via slapadd -q, back-mdb fails: zimbra@zre-ldap001:~$ time /opt/zimbra/openldap/sbin/slapadd -q -b "" -l /tmp/ldap.bak -F /opt/zimbra/data/ldap/config -## 13.49% eta 51m39s elapsed 08m03s spd 1.2 M/s 4f7dee09 mdb_id2entry_put: mdb_put failed: No space left on device(28) "uid=yyyyyyy,ou=people,dc=xxxxxx,dc=com" 4f7dee09 => mdb_tool_entry_put: id2entry_add failed: err=80 4f7dee09 => mdb_tool_entry_put: txn_aborted! Internal error (80) slapadd: could not add entry dn="uid=yyyyyy,ou=people,dc=xxxxx,dc=com" (line=17539364): txn_aborted! Internal error (80) .## 13.50% eta 51m39s elapsed 08m03s spd 1.5 M/s Closing DB...Error, entries missing! entry 20241: uid=zzzzzz,ou=people,dc=xxx,dc=com entry 83243: uid=zzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 85232: uid=zzzzzzzzzzzz,ou=people,dc=xxxxx,dc=com entry 93399: uid=zzzzzzzzzzzzz,ou=people,dc=xxxxx,dc=com entry 113549: uid=zzzzzzzzz,ou=people,dc=xxxxx,dc=com entry 138089: uid=zzzzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 140319: uid=zzzzzzzz,ou=people,dc=xxxxx,dc=com entry 152722: uid=zzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 154989: uid=zzzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 164912: uid=zzzzzzzzz,ou=people,dc=xxxxx,dc=net entry 166738: uid=zzzzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 167049: uid=zzzzzzzzzzz,ou=people,dc=xxxxxxxxxx,dc=com entry 177778: uid=zzzzzzzz,ou=people,dc=xxxxxx,dc=com entry 195907: uid=zzz,ou=people,dc=xxxxxx,dc=com (etc)

1 0

RE: (ITS#7228) Authentication Problem When Using PPolicy and Chaining
by Jong.Limb＠dss.virginia.gov 05 Apr '12

05 Apr '12

I have built and tested 2.4.30, and it works as expected. I now get invalid credentials when I hit the consumer with an incorrect password. However, I believe that the script should still be modified to check the return code from the first ldapsearch in order to prevent something like this developing in the future. You can close this ITS. Thanks. Jong Limb Division of Information Systems Virginia Department of Social Services 804-726-7823 -----Original Message----- From: Limb, Jong (VDSS) Sent: Thursday, April 05, 2012 11:31 AM To: 'Howard Chu' Cc: openldap-its(a)OpenLDAP.org Subject: RE: (ITS#7228) Authentication Problem When Using PPolicy and Chaining I believe the test is incorrect. Near the bottom of the script, there is this section of code: $LDAPSEARCH -H $URI2 -D "$USER" -w wrongpw >$SEARCHOUT 2>&1 $LDAPSEARCH -H $URI1 -D "$MANAGERDN" -w $PASSWD -b "$USER" \* \+ >> $SEARCHOUT 2>&1 COUNT=`grep "pwdFailureTime" $SEARCHOUT | wc -l` echo "jkl 13" if test $COUNT != 1 ; then echo "Policy state forwarding failed" test $KILLSERVERS != no && kill -HUP $KILLPIDS exit 1 fi The first LDAPSEARCH should have its return code checked, and for an incorrect password, it should be 49. If the script is modified to check for 49 and fail otherwise, it will fail. Unfortunately, I am not in a position to use a newer version of OpenLDAP, but I will build it and run the tests to see if the problem exists there as well. I will report back with my findings. Jong Limb Division of Information Systems Virginia Department of Social Services 804-726-7823 -----Original Message----- From: Howard Chu [mailto:hyc@symas.com] Sent: Thursday, April 05, 2012 10:17 AM To: Limb, Jong (VDSS) Cc: openldap-its(a)OpenLDAP.org Subject: Re: (ITS#7228) Authentication Problem When Using PPolicy and Chaining jong.limb(a)dss.virginia.gov wrote: > Full_Name: Jong K. Limb > Version: 2.4.23 > OS: RHEL 5.3 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (166.67.66.5) > > > I have the following setup: > - Provider LDAP server used only by those applications that change passwords > - Consumer LDAP servers used by all applications that want to authenticate > users > - The provider and consumers have password policy configured so that the > accounts lock after 3 failed attempts > - In order to maintain synchronization of failed login attempts across all > consumers, I also enabled chaining from the consumers to the master > > If an attempt is made to authenticate against the consumer with an invalid > password (for example, using ldapsearch), the pwdFailureTime attribute is > added/updated on the provider and eventually synced to the consumer, but the > operation succeeds when it should not have. > > If I remove the password policy overlay (leaving all other configuration the > same) and run ldapsearch with an invalid password against a consumer, the > operation will fail with invalid credentials as it should. > > I have done a little debugging, and it looks like the response that gets > returned to the client is the response to the modify operation that the consumer > makes to the provider to add/update the pwdFailureTime attribute, and not the > response to the bind operation. The modify operation is successful here, so the > client continues on with the search or other operation. Unable to confirm this. Note that this is tested explicitly in test022 of the test suite, and no such behavior occurs there. You're running a pretty old release, perhaps you should update. Regardless, unless you can provide more details to reproduce this situation, this ITS will be closed. -- -- Howard Chu CTO, 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#7228) Authentication Problem When Using PPolicy and Chaining
by Jong.Limb＠dss.virginia.gov 05 Apr '12

05 Apr '12

I believe the test is incorrect. Near the bottom of the script, there is this section of code: $LDAPSEARCH -H $URI2 -D "$USER" -w wrongpw >$SEARCHOUT 2>&1 $LDAPSEARCH -H $URI1 -D "$MANAGERDN" -w $PASSWD -b "$USER" \* \+ >> $SEARCHOUT 2>&1 COUNT=`grep "pwdFailureTime" $SEARCHOUT | wc -l` echo "jkl 13" if test $COUNT != 1 ; then echo "Policy state forwarding failed" test $KILLSERVERS != no && kill -HUP $KILLPIDS exit 1 fi The first LDAPSEARCH should have its return code checked, and for an incorrect password, it should be 49. If the script is modified to check for 49 and fail otherwise, it will fail. Unfortunately, I am not in a position to use a newer version of OpenLDAP, but I will build it and run the tests to see if the problem exists there as well. I will report back with my findings. Jong Limb Division of Information Systems Virginia Department of Social Services 804-726-7823 -----Original Message----- From: Howard Chu [mailto:hyc@symas.com] Sent: Thursday, April 05, 2012 10:17 AM To: Limb, Jong (VDSS) Cc: openldap-its(a)OpenLDAP.org Subject: Re: (ITS#7228) Authentication Problem When Using PPolicy and Chaining jong.limb(a)dss.virginia.gov wrote: > Full_Name: Jong K. Limb > Version: 2.4.23 > OS: RHEL 5.3 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (166.67.66.5) > > > I have the following setup: > - Provider LDAP server used only by those applications that change passwords > - Consumer LDAP servers used by all applications that want to authenticate > users > - The provider and consumers have password policy configured so that the > accounts lock after 3 failed attempts > - In order to maintain synchronization of failed login attempts across all > consumers, I also enabled chaining from the consumers to the master > > If an attempt is made to authenticate against the consumer with an invalid > password (for example, using ldapsearch), the pwdFailureTime attribute is > added/updated on the provider and eventually synced to the consumer, but the > operation succeeds when it should not have. > > If I remove the password policy overlay (leaving all other configuration the > same) and run ldapsearch with an invalid password against a consumer, the > operation will fail with invalid credentials as it should. > > I have done a little debugging, and it looks like the response that gets > returned to the client is the response to the modify operation that the consumer > makes to the provider to add/update the pwdFailureTime attribute, and not the > response to the bind operation. The modify operation is successful here, so the > client continues on with the search or other operation. Unable to confirm this. Note that this is tested explicitly in test022 of the test suite, and no such behavior occurs there. You're running a pretty old release, perhaps you should update. Regardless, unless you can provide more details to reproduce this situation, this ITS will be closed. -- -- Howard Chu CTO, 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#7226) olcAuditlogFile At dos not accept multiple values
by michael＠stroeder.com 05 Apr '12

05 Apr '12

This is a cryptographically signed message in MIME format. --------------ms000601060706020001090605 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable hyc(a)symas.com wrote: > Michael Str=F6der wrote: >> hyc(a)symas.com wrote: >>> michael(a)stroeder.com wrote: >>>> 2. add an equality matching rule. >>> >>> My approach with config attrs has been to skip the equality rule for >>> single-valued attrs. Since they are single-valued, there's obviously = no need >>> to allow for comparing multiple values. >> >> They are not searchable either then. Does it add too much overhead to = add an >> equality matching rule? >> >> Personally I currently can't imagine a use-case for searching an entry= using >> attribute olcAuditlogFile but one never knows... >=20 > True, depends on the syntax and whether it requires a normalizer, and I= =20 > couldn't think of a use-case for searching either. Overall if there's n= o=20 > obvious use for it, I choose to omit it. Hmm, I come to the opposite conclusion. Why? When generating the diff between old and new entry my web2ldap looks whet= her an EQUALITY matching rule is defined for a certain attribute type and it'= s also published in attributes matchingRules in the subschema subentry. In = that case old attribute values are explicitly deleted (not just the whole attr= ibute). This is done to provoke collisions on the server's side during processing= the modify request to be more on the safe side in case different users alter = the same entries. This has also been discussed some time ago on the OpenLDAP = lists. (Additionally the assertion control with appopriate filter is sent along with the modify request for that purpose. But not if the server is OpenLDAP because of ITS#6916 which renders assertion control completely u= nusable.) =3D> If there's no objection against an equality rule (e.g. because of overhead) *every* attribute type which uses a LDAP syntax for which an equality rule exists should have one. Ciao, Michael. --------------ms000601060706020001090605 Content-Type: application/pkcs7-signature; name="smime.p7s" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="smime.p7s" Content-Description: S/MIME Cryptographic Signature MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAQAAoIIFOzCC BTcwggMfoAMCAQICAwl4kDANBgkqhkiG9w0BAQUFADB5MRAwDgYDVQQKEwdSb290IENBMR4w HAYDVQQLExVodHRwOi8vd3d3LmNhY2VydC5vcmcxIjAgBgNVBAMTGUNBIENlcnQgU2lnbmlu ZyBBdXRob3JpdHkxITAfBgkqhkiG9w0BCQEWEnN1cHBvcnRAY2FjZXJ0Lm9yZzAeFw0xMDEx MTcxNTQzMzFaFw0xMjExMTYxNTQzMzFaMD8xGDAWBgNVBAMUD01pY2hhZWwgU3Ry9mRlcjEj MCEGCSqGSIb3DQEJARYUbWljaGFlbEBzdHJvZWRlci5jb20wggEiMA0GCSqGSIb3DQEBAQUA A4IBDwAwggEKAoIBAQDo2SKth5GhtaDrCyfGtyUG+/hAAa/J52L0NFN4SSRvTtdGf9HfWwwd NCtgae0TVGWk2lKDbXA9d5vmyIiRhuwxd90H6FLErhRBeB9G67qtw87E8WUoXt2DwPQEUTWV hqHpPadlmgFw3+i3TGQQTe3O3W9MMMd4GJNhObem2VGRuCD37OXnzBksTcq0FPJgcWAhe3d/ 0ItOkNWBqgq8Mf3p7WFBhaQ0a27BC/mKtH8fI3kPcS305imPRja69Msq3EwUZBc9ToVp6FRQ NYKjfOBybDUzVkmRZl3H8xutQP2w8Zxb8m5f7Q1BfLLrIFScfYvIDgOERxTCd4lab8+/09XH AgMBAAGjggEAMIH9MAwGA1UdEwEB/wQCMAAwVgYJYIZIAYb4QgENBEkWR1RvIGdldCB5b3Vy IG93biBjZXJ0aWZpY2F0ZSBmb3IgRlJFRSBoZWFkIG92ZXIgdG8gaHR0cDovL3d3dy5DQWNl cnQub3JnMEAGA1UdJQQ5MDcGCCsGAQUFBwMEBggrBgEFBQcDAgYKKwYBBAGCNwoDBAYKKwYB BAGCNwoDAwYJYIZIAYb4QgQBMDIGCCsGAQUFBwEBBCYwJDAiBggrBgEFBQcwAYYWaHR0cDov L29jc3AuY2FjZXJ0Lm9yZzAfBgNVHREEGDAWgRRtaWNoYWVsQHN0cm9lZGVyLmNvbTANBgkq hkiG9w0BAQUFAAOCAgEANPf/aLF41eQlvN5dEg3CFnlN//qQK7+EPIXLnHprNWLb4nBwgdPj /E+qa1umT7px4Py3VS0UTKqLmMdWftwid8MOMHWalZwrfx0Z8U3He+EdJhOSnn9vdd/ug7Xd dI/hRjLaBSq9ZhCczEUgL6vTxCYPlIoHF56y/oxSJw59vRBjvRFKXvpBZWseeRkcGACQduNH SNdWC1IqHAbQlgOS9VWQUYlm//BdaLkezRxqnQp5+KJMAcZzHpdNJ3G4SqCJ02Z3n4kk8IKZ AjgiWxisDFNsfXKDb9Ng5ntnnH2ouxrgPoNnW445tgkz50VKHstylx9s5O3G7uUTtg0J+z63 TA8xbN6kzRx7RgAUkEXhl6WEdW+3EVj5tYY38Uy8vleP+gYZfphKEmQJgIQqy9D2+gesbolT QdWYgbUYY2AHJOshskMW7pahYnFX2pZmn/ayaPc+JFJlCEqO0+DcYQjYuv6sntQgZGkok7yZ R4xMbyCp61pTrfGWOufZs/FiScJZg1IWY5qb4URH4VZZjLNMR2pFMRuE4LvgkkMRasbUv7Yv n3Lzv34lTfJKUqYW6nx//L2NS4rN63o0taPwRygnuBK4kp7EYEcwtLeanJhQoIu4b6If9rwy D7CFAp51wIewV9VtZ1Is0irNBcMVyhJogIcuIn+VWY1ff1RxySD/djMxggOUMIIDkAIBATCB gDB5MRAwDgYDVQQKEwdSb290IENBMR4wHAYDVQQLExVodHRwOi8vd3d3LmNhY2VydC5vcmcx IjAgBgNVBAMTGUNBIENlcnQgU2lnbmluZyBBdXRob3JpdHkxITAfBgkqhkiG9w0BCQEWEnN1 cHBvcnRAY2FjZXJ0Lm9yZwIDCXiQMAkGBSsOAwIaBQCgggHoMBgGCSqGSIb3DQEJAzELBgkq hkiG9w0BBwEwHAYJKoZIhvcNAQkFMQ8XDTEyMDQwNTE0MjAyOVowIwYJKoZIhvcNAQkEMRYE FFzAsG3+dnjGgCIuHe9WJ6JWCSCgMF8GCSqGSIb3DQEJDzFSMFAwCwYJYIZIAWUDBAECMAoG CCqGSIb3DQMHMA4GCCqGSIb3DQMCAgIAgDANBggqhkiG9w0DAgIBQDAHBgUrDgMCBzANBggq hkiG9w0DAgIBKDCBkQYJKwYBBAGCNxAEMYGDMIGAMHkxEDAOBgNVBAoTB1Jvb3QgQ0ExHjAc BgNVBAsTFWh0dHA6Ly93d3cuY2FjZXJ0Lm9yZzEiMCAGA1UEAxMZQ0EgQ2VydCBTaWduaW5n IEF1dGhvcml0eTEhMB8GCSqGSIb3DQEJARYSc3VwcG9ydEBjYWNlcnQub3JnAgMJeJAwgZMG CyqGSIb3DQEJEAILMYGDoIGAMHkxEDAOBgNVBAoTB1Jvb3QgQ0ExHjAcBgNVBAsTFWh0dHA6 Ly93d3cuY2FjZXJ0Lm9yZzEiMCAGA1UEAxMZQ0EgQ2VydCBTaWduaW5nIEF1dGhvcml0eTEh MB8GCSqGSIb3DQEJARYSc3VwcG9ydEBjYWNlcnQub3JnAgMJeJAwDQYJKoZIhvcNAQEBBQAE ggEA2W9KZ+2wjxLM0svAhmhXXcOJGksGQI+N0u4ta2/ddnRZwYmlX2ff0OeC0XRom5x6dmIV SzO9Hiw2syv7FRq7dIEI+lz0XWC1OttnG5QQTxyLR52EW/0kXt+eP+HEIcNqMbeZAC8JPKmr XEnpOC1ZwhxZYx2Gq72V379rkohDiW8hWuJdnqKcxFnlpzYd/XtmOhhnOLhOUKF7mYiqnTuz dhPETknUf+0rLjGCBMXVuZ94kWbU3Ia4gxtwVbRfruhadNBO61Dx3zccWqIdmBjzdNkFmIpo i4pjcwfRzgyagZxcazpOihioFxEu8bA3PQN8PUgAlvMdbU9RHrmwICrUPwAAAAAAAA== --------------ms000601060706020001090605--

1 0

Re: (ITS#7228) Authentication Problem When Using PPolicy and Chaining
by hyc＠symas.com 05 Apr '12

05 Apr '12

jong.limb(a)dss.virginia.gov wrote: > Full_Name: Jong K. Limb > Version: 2.4.23 > OS: RHEL 5.3 > URL: ftp://ftp.openldap.org/incoming/ > Submission from: (NULL) (166.67.66.5) > > > I have the following setup: > - Provider LDAP server used only by those applications that change passwords > - Consumer LDAP servers used by all applications that want to authenticate > users > - The provider and consumers have password policy configured so that the > accounts lock after 3 failed attempts > - In order to maintain synchronization of failed login attempts across all > consumers, I also enabled chaining from the consumers to the master > > If an attempt is made to authenticate against the consumer with an invalid > password (for example, using ldapsearch), the pwdFailureTime attribute is > added/updated on the provider and eventually synced to the consumer, but the > operation succeeds when it should not have. > > If I remove the password policy overlay (leaving all other configuration the > same) and run ldapsearch with an invalid password against a consumer, the > operation will fail with invalid credentials as it should. > > I have done a little debugging, and it looks like the response that gets > returned to the client is the response to the modify operation that the consumer > makes to the provider to add/update the pwdFailureTime attribute, and not the > response to the bind operation. The modify operation is successful here, so the > client continues on with the search or other operation. Unable to confirm this. Note that this is tested explicitly in test022 of the test suite, and no such behavior occurs there. You're running a pretty old release, perhaps you should update. Regardless, unless you can provide more details to reproduce this situation, this ITS will be closed. -- -- Howard Chu CTO, Symas Corp. http://www.symas.com Director, Highland Sun http://highlandsun.com/hyc/ Chief Architect, OpenLDAP http://www.openldap.org/project/

1 0

(ITS#7228) Authentication Problem When Using PPolicy and Chaining
by jong.limb＠dss.virginia.gov 05 Apr '12

05 Apr '12

Full_Name: Jong K. Limb Version: 2.4.23 OS: RHEL 5.3 URL: ftp://ftp.openldap.org/incoming/ Submission from: (NULL) (166.67.66.5) I have the following setup: - Provider LDAP server used only by those applications that change passwords - Consumer LDAP servers used by all applications that want to authenticate users - The provider and consumers have password policy configured so that the accounts lock after 3 failed attempts - In order to maintain synchronization of failed login attempts across all consumers, I also enabled chaining from the consumers to the master If an attempt is made to authenticate against the consumer with an invalid password (for example, using ldapsearch), the pwdFailureTime attribute is added/updated on the provider and eventually synced to the consumer, but the operation succeeds when it should not have. If I remove the password policy overlay (leaving all other configuration the same) and run ldapsearch with an invalid password against a consumer, the operation will fail with invalid credentials as it should. I have done a little debugging, and it looks like the response that gets returned to the client is the response to the modify operation that the consumer makes to the provider to add/update the pwdFailureTime attribute, and not the response to the bind operation. The modify operation is successful here, so the client continues on with the search or other operation.

1 0

Re: (ITS#7226) olcAuditlogFile At dos not accept multiple values
by hyc＠symas.com 05 Apr '12

05 Apr '12

Michael Ströder wrote: > hyc(a)symas.com wrote: >> michael(a)stroeder.com wrote: >>> 2. add an equality matching rule. >> >> My approach with config attrs has been to skip the equality rule for >> single-valued attrs. Since they are single-valued, there's obviously no need >> to allow for comparing multiple values. > > They are not searchable either then. Does it add too much overhead to add an > equality matching rule? > > Personally I currently can't imagine a use-case for searching an entry using > attribute olcAuditlogFile but one never knows... True, depends on the syntax and whether it requires a normalizer, and I couldn't think of a use-case for searching either. Overall if there's no obvious use for it, I choose to omit it. -- -- Howard Chu CTO, 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#7226) olcAuditlogFile At dos not accept multiple values
by michael＠stroeder.com 05 Apr '12

05 Apr '12

This is a cryptographically signed message in MIME format. --------------ms030905050806080408090400 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable hyc(a)symas.com wrote: > michael(a)stroeder.com wrote: >> 2. add an equality matching rule. >=20 > My approach with config attrs has been to skip the equality rule for=20 > single-valued attrs. Since they are single-valued, there's obviously no= need=20 > to allow for comparing multiple values. They are not searchable either then. Does it add too much overhead to add= an equality matching rule? Personally I currently can't imagine a use-case for searching an entry us= ing attribute olcAuditlogFile but one never knows... Ciao, Michael. --------------ms030905050806080408090400 Content-Type: application/pkcs7-signature; name="smime.p7s" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="smime.p7s" Content-Description: S/MIME Cryptographic Signature MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAQAAoIIFOzCC BTcwggMfoAMCAQICAwl4kDANBgkqhkiG9w0BAQUFADB5MRAwDgYDVQQKEwdSb290IENBMR4w HAYDVQQLExVodHRwOi8vd3d3LmNhY2VydC5vcmcxIjAgBgNVBAMTGUNBIENlcnQgU2lnbmlu ZyBBdXRob3JpdHkxITAfBgkqhkiG9w0BCQEWEnN1cHBvcnRAY2FjZXJ0Lm9yZzAeFw0xMDEx MTcxNTQzMzFaFw0xMjExMTYxNTQzMzFaMD8xGDAWBgNVBAMUD01pY2hhZWwgU3Ry9mRlcjEj MCEGCSqGSIb3DQEJARYUbWljaGFlbEBzdHJvZWRlci5jb20wggEiMA0GCSqGSIb3DQEBAQUA A4IBDwAwggEKAoIBAQDo2SKth5GhtaDrCyfGtyUG+/hAAa/J52L0NFN4SSRvTtdGf9HfWwwd NCtgae0TVGWk2lKDbXA9d5vmyIiRhuwxd90H6FLErhRBeB9G67qtw87E8WUoXt2DwPQEUTWV hqHpPadlmgFw3+i3TGQQTe3O3W9MMMd4GJNhObem2VGRuCD37OXnzBksTcq0FPJgcWAhe3d/ 0ItOkNWBqgq8Mf3p7WFBhaQ0a27BC/mKtH8fI3kPcS305imPRja69Msq3EwUZBc9ToVp6FRQ NYKjfOBybDUzVkmRZl3H8xutQP2w8Zxb8m5f7Q1BfLLrIFScfYvIDgOERxTCd4lab8+/09XH AgMBAAGjggEAMIH9MAwGA1UdEwEB/wQCMAAwVgYJYIZIAYb4QgENBEkWR1RvIGdldCB5b3Vy IG93biBjZXJ0aWZpY2F0ZSBmb3IgRlJFRSBoZWFkIG92ZXIgdG8gaHR0cDovL3d3dy5DQWNl cnQub3JnMEAGA1UdJQQ5MDcGCCsGAQUFBwMEBggrBgEFBQcDAgYKKwYBBAGCNwoDBAYKKwYB BAGCNwoDAwYJYIZIAYb4QgQBMDIGCCsGAQUFBwEBBCYwJDAiBggrBgEFBQcwAYYWaHR0cDov L29jc3AuY2FjZXJ0Lm9yZzAfBgNVHREEGDAWgRRtaWNoYWVsQHN0cm9lZGVyLmNvbTANBgkq hkiG9w0BAQUFAAOCAgEANPf/aLF41eQlvN5dEg3CFnlN//qQK7+EPIXLnHprNWLb4nBwgdPj /E+qa1umT7px4Py3VS0UTKqLmMdWftwid8MOMHWalZwrfx0Z8U3He+EdJhOSnn9vdd/ug7Xd dI/hRjLaBSq9ZhCczEUgL6vTxCYPlIoHF56y/oxSJw59vRBjvRFKXvpBZWseeRkcGACQduNH SNdWC1IqHAbQlgOS9VWQUYlm//BdaLkezRxqnQp5+KJMAcZzHpdNJ3G4SqCJ02Z3n4kk8IKZ AjgiWxisDFNsfXKDb9Ng5ntnnH2ouxrgPoNnW445tgkz50VKHstylx9s5O3G7uUTtg0J+z63 TA8xbN6kzRx7RgAUkEXhl6WEdW+3EVj5tYY38Uy8vleP+gYZfphKEmQJgIQqy9D2+gesbolT QdWYgbUYY2AHJOshskMW7pahYnFX2pZmn/ayaPc+JFJlCEqO0+DcYQjYuv6sntQgZGkok7yZ R4xMbyCp61pTrfGWOufZs/FiScJZg1IWY5qb4URH4VZZjLNMR2pFMRuE4LvgkkMRasbUv7Yv n3Lzv34lTfJKUqYW6nx//L2NS4rN63o0taPwRygnuBK4kp7EYEcwtLeanJhQoIu4b6If9rwy D7CFAp51wIewV9VtZ1Is0irNBcMVyhJogIcuIn+VWY1ff1RxySD/djMxggOUMIIDkAIBATCB gDB5MRAwDgYDVQQKEwdSb290IENBMR4wHAYDVQQLExVodHRwOi8vd3d3LmNhY2VydC5vcmcx IjAgBgNVBAMTGUNBIENlcnQgU2lnbmluZyBBdXRob3JpdHkxITAfBgkqhkiG9w0BCQEWEnN1 cHBvcnRAY2FjZXJ0Lm9yZwIDCXiQMAkGBSsOAwIaBQCgggHoMBgGCSqGSIb3DQEJAzELBgkq hkiG9w0BBwEwHAYJKoZIhvcNAQkFMQ8XDTEyMDQwNTEzMTE1MlowIwYJKoZIhvcNAQkEMRYE FKSAdbqqyHLXa1gjxrZTVmU3u7tyMF8GCSqGSIb3DQEJDzFSMFAwCwYJYIZIAWUDBAECMAoG CCqGSIb3DQMHMA4GCCqGSIb3DQMCAgIAgDANBggqhkiG9w0DAgIBQDAHBgUrDgMCBzANBggq hkiG9w0DAgIBKDCBkQYJKwYBBAGCNxAEMYGDMIGAMHkxEDAOBgNVBAoTB1Jvb3QgQ0ExHjAc BgNVBAsTFWh0dHA6Ly93d3cuY2FjZXJ0Lm9yZzEiMCAGA1UEAxMZQ0EgQ2VydCBTaWduaW5n IEF1dGhvcml0eTEhMB8GCSqGSIb3DQEJARYSc3VwcG9ydEBjYWNlcnQub3JnAgMJeJAwgZMG CyqGSIb3DQEJEAILMYGDoIGAMHkxEDAOBgNVBAoTB1Jvb3QgQ0ExHjAcBgNVBAsTFWh0dHA6 Ly93d3cuY2FjZXJ0Lm9yZzEiMCAGA1UEAxMZQ0EgQ2VydCBTaWduaW5nIEF1dGhvcml0eTEh MB8GCSqGSIb3DQEJARYSc3VwcG9ydEBjYWNlcnQub3JnAgMJeJAwDQYJKoZIhvcNAQEBBQAE ggEA19nxcp9VtMesBkigrXCwA3bb5d6mDGvRHxlHY98mpA/uVI5tjiEgya9H2Eby+z34Ws7o UtaSyQ4WbDT2SmtuGt7SKd9+DqpBRQnsOu52rXo31vBoQ+RCASOfix72v8geh3Cn6Ft5ZDX4 lZ2qYc8UmEYiY04nLkW4cIvCChd5eREiPBECxHhcrAXt/Rv4JNUhqOsQ90uFZvy6d90ZMHMZ kRtBF7O36gV9Wu/K6Oa3yWrOIdDyXeWIpIeIh2UrU3ZPDjeiRvtB0iHgmbhnOLEeWMMtA9bU XmQdTqL977WN4Ga3xY952sNExcEmoHMcvgqFeFoLvp0hfrCEbGFTVmmH0wAAAAAAAA== --------------ms030905050806080408090400--

1 0

Re: (ITS#7226) olcAuditlogFile At dos not accept multiple values
by hyc＠symas.com 05 Apr '12

05 Apr '12

michael(a)stroeder.com wrote: > Emmanuel Lécharny wrote: >> One more question : why is this AT multi-valued ? Does it make sense ? (Just >> curious) > > I tried this static configuration: > > overlay auditlog > auditlog /tmp/auditlog1.ldif > auditlog /tmp/auditlog2.ldif > > Only file /tmp/auditlog2.ldif is created and changes get writte to this single > file. > > So yes, I support to > 1. set SINGLE-VALUE for this attribute type and Probably a good idea. > 2. add an equality matching rule. My approach with config attrs has been to skip the equality rule for single-valued attrs. Since they are single-valued, there's obviously no need to allow for comparing multiple values. -- -- Howard Chu CTO, 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:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

openldap-bugs April 2012