Re: [Freeipa-users] Windows sync agreement becomes uninitialized and crashes directory server

2015-07-13 Thread Rich Megginson

On 07/13/2015 07:07 PM, nat...@nathanpeters.com wrote:

2 FreeIPA 4.1.4 servers running on CentOS 7.
dc1 has a sync agreement to a windows server.


It has been running fine since June 5 when I re-initialized a sync
agreement that had somehow uninitialized itself.  Original issue report
here :
https://www.redhat.com/archives/freeipa-users/2015-June/msg00147.html
Bug report here : https://fedorahosted.org/freeipa/ticket/5054

It appears the same thing may have happened again, one month later, but
this time randomly, as we have not made any changes to our sync agreement
since the initial change in June.  it appears to have unitialized itself
without us changing it and managed to crash the directory server in doing
so.


Crash? http://www.port389.org/docs/389ds/FAQ/faq.html#debugging-crashes
# debuginfo-install 389-ds-base ipa-server slapi-nis


  Note that during the last week I could still login to the web ui, but
around the time the log entries change, I became unable to.

I tried to login to the web server today and it would not let me login, so
I went to the shell on the server and noticed that ipactl command would
freeze up again.  I looked at the logs (which I will paste below) and
restarted the directory server service.

I then found that my sync agreement had become uninitialized.

--- output ---
[root@dc1 slapd-IPADOMAIN-NET]# ldapsearch -xLLL -D cn=directory manager
-W -b cn=config objectclass=nsDSWindowsReplicationAgreement
Enter LDAP Password:
dn: cn=meToofficedc2.office.addomain.net,cn=replica,cn=dc\3Dipadomain
  \2Cdc\3Dnet,cn=mapping tree,cn=config
nsds7WindowsReplicaSubtree: OU=Staff,DC=office,DC=addomain,DC=net
nsds7DirectoryReplicaSubtree: cn=users,cn=accounts,dc=ipadomain,dc=net
cn: meToofficedc2.office.addomain.net
nsds7NewWinGroupSyncEnabled: false
objectClass: nsDSWindowsReplicationAgreement
objectClass: top
nsDS5ReplicaTransportInfo: TLS
description: me to officedc2.office.addomain.net
nsDS5ReplicaRoot: dc=ipadomain,dc=net
nsDS5ReplicaHost: officedc2.office.addomain.net
nsds5replicaTimeout: 120
nsDS5ReplicaBindDN: cn=freeipa syncuser,ou=Service
Account,dc=office,dc=addomain,dc=net
nsds7NewWinUserSyncEnabled: true
nsDS5ReplicaPort: 389
nsds7WindowsDomain: ipadomain.net
nsDS5ReplicatedAttributeList: (objectclass=*) $ EXCLUDE memberof
idnssoaserial
   entryusn krblastsuccessfulauth krblastfailedauth krbloginfailedcount
nsDS5ReplicaBindMethod: simple
nsDS5ReplicaCredentials:
{AES-TUhNR0NTcUdTSWIzRFFFRkRUQm1NRVVHQ1NxR1NJYjNEUUVG
  RERBNEJDUmtOelUzTTJJNVlpMDBaV1EyTTJRMQ0KWXkwNU0yTm1aV05sTVMxbU5qRXpaak5oTlFBQ
  0FRSUNBU0F3Q2dZSUtvWklodmNOQWdjd0hRWUpZSVpJQVdVRA0KQkFFcUJCQ2k0N0NxRGZFd2JIdm
  I0MFVFZVI3MA==}gWI9NIB8lbt9tmNszzbBFCAe4Vs/e0sMyn5+NZPJg9E=
nsds7DirsyncCookie::
TVNEUwMAAABoJPGME7jQAQAAYAEAAPc1qQAAA
  AD3NakAAMUjuImqVZhBkOkdt24C0IsBAA4AY4GwFkVcvEmMMExrVon4d6
  13PwAADGzFNzznrESIxHzA74fbsz4HUgAAOnFoO5OE2E27lR/g4EcjQTLbIwAAuEm
  PWjYok0qGS0HM/+TDmK7FgAMA6PTFXvAdnkaJSIkZT1lS+/FcJAAA4qTQaC46/Ua4KXgP
  /ixNcbK4dgAAWowbgYD1akibZ+sCul5C4VNsMQAAxSO4iapVmEGQ6R23bgLQi/c1qQAAA
  AAAogC6jFcyFUmhBp4B7FkaBcRHwwEAyhKMxsP0uUKGEnG2lsyA8eTUwgYA4n8Xx1bAlU
  mBUl3zhlZ9WBngDAAA71vM2ebFEkCJkBaLjB4CGU+4CQMAGfO+4ndZCkaVKnwZNlNsf90
  NDAAAgD6n+M2bcUGkOwo5gPLx7IOjAwAA
nsds50ruv: {replicageneration} 553fe9bb0004
nsds50ruv: {replica 4 ldap://dc1.ipadomain.net:389} 553fe9c9
  0004 557f49fb00020004
nsds50ruv: {replica 3 ldap://dc2.ipadomain.net:389} 553fe9c
  40003 557f3e4a00020003
nsruvReplicaLastModified: {replica 4 ldap://dc1.ipadomain.ne
  t:389} 557f494a
nsruvReplicaLastModified: {replica 3 ldap://dc2.ipadomain.n
  et:389} 557f3d95
oneWaySync: fromWindows
nsds5ReplicaEnabled: on
nsds5replicareapactive: 0
nsds5replicaLastUpdateStart: 0
nsds5replicaLastUpdateEnd: 0
nsds5replicaChangesSentSinceStartup:
nsds5replicaLastUpdateStatus: -1  - LDAP error: Can't contact LDAP server
nsds5replicaUpdateInProgress: FALSE
nsds5replicaLastInitStart: 0
nsds5replicaLastInitEnd: 0
--- output ---

Here are the error logs for the last month for the directory server.  They
are totally empty until July 2.

---output---
 389-Directory/1.3.3.8 B2015.040.128
 dc1.ipadomain.net:636 (/etc/dirsrv/slapd-IPADOMAIN-NET)

[02/Jul/2015:03:19:02 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[02/Jul/2015:06:10:29 +] - Entry
uid=jenkinsdev,cn=users,cn=accounts,dc=ipadomain,dc=net missing
attribute sn required by object class person
[03/Jul/2015:02:04:02 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:05:39:01 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:17:09:00 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:22:41:32 +] slapd_ldap_sasl_interactive_bind - Error:
could not perform interactive bind for id 

[Freeipa-users] Windows sync agreement becomes uninitialized and crashes directory server

2015-07-13 Thread nathan
2 FreeIPA 4.1.4 servers running on CentOS 7.
dc1 has a sync agreement to a windows server.


It has been running fine since June 5 when I re-initialized a sync
agreement that had somehow uninitialized itself.  Original issue report
here :
https://www.redhat.com/archives/freeipa-users/2015-June/msg00147.html
Bug report here : https://fedorahosted.org/freeipa/ticket/5054

It appears the same thing may have happened again, one month later, but
this time randomly, as we have not made any changes to our sync agreement
since the initial change in June.  it appears to have unitialized itself
without us changing it and managed to crash the directory server in doing
so.  Note that during the last week I could still login to the web ui, but
around the time the log entries change, I became unable to.

I tried to login to the web server today and it would not let me login, so
I went to the shell on the server and noticed that ipactl command would
freeze up again.  I looked at the logs (which I will paste below) and
restarted the directory server service.

I then found that my sync agreement had become uninitialized.

--- output ---
[root@dc1 slapd-IPADOMAIN-NET]# ldapsearch -xLLL -D cn=directory manager
-W -b cn=config objectclass=nsDSWindowsReplicationAgreement
Enter LDAP Password:
dn: cn=meToofficedc2.office.addomain.net,cn=replica,cn=dc\3Dipadomain
 \2Cdc\3Dnet,cn=mapping tree,cn=config
nsds7WindowsReplicaSubtree: OU=Staff,DC=office,DC=addomain,DC=net
nsds7DirectoryReplicaSubtree: cn=users,cn=accounts,dc=ipadomain,dc=net
cn: meToofficedc2.office.addomain.net
nsds7NewWinGroupSyncEnabled: false
objectClass: nsDSWindowsReplicationAgreement
objectClass: top
nsDS5ReplicaTransportInfo: TLS
description: me to officedc2.office.addomain.net
nsDS5ReplicaRoot: dc=ipadomain,dc=net
nsDS5ReplicaHost: officedc2.office.addomain.net
nsds5replicaTimeout: 120
nsDS5ReplicaBindDN: cn=freeipa syncuser,ou=Service
Account,dc=office,dc=addomain,dc=net
nsds7NewWinUserSyncEnabled: true
nsDS5ReplicaPort: 389
nsds7WindowsDomain: ipadomain.net
nsDS5ReplicatedAttributeList: (objectclass=*) $ EXCLUDE memberof
idnssoaserial
  entryusn krblastsuccessfulauth krblastfailedauth krbloginfailedcount
nsDS5ReplicaBindMethod: simple
nsDS5ReplicaCredentials:
{AES-TUhNR0NTcUdTSWIzRFFFRkRUQm1NRVVHQ1NxR1NJYjNEUUVG
 RERBNEJDUmtOelUzTTJJNVlpMDBaV1EyTTJRMQ0KWXkwNU0yTm1aV05sTVMxbU5qRXpaak5oTlFBQ
 0FRSUNBU0F3Q2dZSUtvWklodmNOQWdjd0hRWUpZSVpJQVdVRA0KQkFFcUJCQ2k0N0NxRGZFd2JIdm
 I0MFVFZVI3MA==}gWI9NIB8lbt9tmNszzbBFCAe4Vs/e0sMyn5+NZPJg9E=
nsds7DirsyncCookie::
TVNEUwMAAABoJPGME7jQAQAAYAEAAPc1qQAAA
 AD3NakAAMUjuImqVZhBkOkdt24C0IsBAA4AY4GwFkVcvEmMMExrVon4d6
 13PwAADGzFNzznrESIxHzA74fbsz4HUgAAOnFoO5OE2E27lR/g4EcjQTLbIwAAuEm
 PWjYok0qGS0HM/+TDmK7FgAMA6PTFXvAdnkaJSIkZT1lS+/FcJAAA4qTQaC46/Ua4KXgP
 /ixNcbK4dgAAWowbgYD1akibZ+sCul5C4VNsMQAAxSO4iapVmEGQ6R23bgLQi/c1qQAAA
 AAAogC6jFcyFUmhBp4B7FkaBcRHwwEAyhKMxsP0uUKGEnG2lsyA8eTUwgYA4n8Xx1bAlU
 mBUl3zhlZ9WBngDAAA71vM2ebFEkCJkBaLjB4CGU+4CQMAGfO+4ndZCkaVKnwZNlNsf90
 NDAAAgD6n+M2bcUGkOwo5gPLx7IOjAwAA
nsds50ruv: {replicageneration} 553fe9bb0004
nsds50ruv: {replica 4 ldap://dc1.ipadomain.net:389} 553fe9c9
 0004 557f49fb00020004
nsds50ruv: {replica 3 ldap://dc2.ipadomain.net:389} 553fe9c
 40003 557f3e4a00020003
nsruvReplicaLastModified: {replica 4 ldap://dc1.ipadomain.ne
 t:389} 557f494a
nsruvReplicaLastModified: {replica 3 ldap://dc2.ipadomain.n
 et:389} 557f3d95
oneWaySync: fromWindows
nsds5ReplicaEnabled: on
nsds5replicareapactive: 0
nsds5replicaLastUpdateStart: 0
nsds5replicaLastUpdateEnd: 0
nsds5replicaChangesSentSinceStartup:
nsds5replicaLastUpdateStatus: -1  - LDAP error: Can't contact LDAP server
nsds5replicaUpdateInProgress: FALSE
nsds5replicaLastInitStart: 0
nsds5replicaLastInitEnd: 0
--- output ---

Here are the error logs for the last month for the directory server.  They
are totally empty until July 2.

---output---
389-Directory/1.3.3.8 B2015.040.128
dc1.ipadomain.net:636 (/etc/dirsrv/slapd-IPADOMAIN-NET)

[02/Jul/2015:03:19:02 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[02/Jul/2015:06:10:29 +] - Entry
uid=jenkinsdev,cn=users,cn=accounts,dc=ipadomain,dc=net missing
attribute sn required by object class person
[03/Jul/2015:02:04:02 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:05:39:01 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:17:09:00 +] NSMMReplicationPlugin - windows sync - failed
to send dirsync search request: 2
[03/Jul/2015:22:41:32 +] slapd_ldap_sasl_interactive_bind - Error:
could not perform interactive bind for id [] mech [GSSAPI]: LDAP error -2
(Local error) (SASL(-1): generic failure: GSSAPI Error: Unspecified GSS
failure.  Minor code may provide more information (Cannot contact any KDC
for realm