View Issue Details

IDProjectCategoryView StatusLast Update
0015860CentOS-7sssdpublic2019-08-20 16:48
ReporterHenrik 
PrioritynormalSeveritytrivialReproducibilityN/A
Status newResolutionopen 
Product Version7.5.1804 
Target VersionFixed in Version 
Summary0015860: Update sssd produce segfault - error 5 in libtalloc
DescriptionHello,

I have lot of Linux CentOS VMs and got a issue with one that colleagues couldn't login to a specific VM anymore. Systemctl sssd as also /var/log messages we're spamming following error messages:

"systemctl status sssd" and /var/log/messages was spamming following error:
"[sssd[ldap_child[11108]]]: Failed to initialize credentials using keytab [MEMORY:/etc/krb5.keytab]: Preauthentication failed. Unable to create GSSAPI-encrypted LDAP connection."

- ntp force update didn't help
- Realm leave / realm join didn't help
- systemctl restart sssd was possible and came up with status "active (running)" but spammed again "GSSAPI-encrypted" errors


 
Steps To ReproduceAs steps like "ntp / sssd restarts / realm join-leave" didn't solve the issue. I performed following steps

1. yum check-update
2. yum update sssd
3. Check realm join/leave if error is gone and login would be possible again.

3.1 sssd startet to segfault with "error 5 in libtalloc.so".
"kernel: sssd_be[7187]: segfault at ffffffffffffffa1 ip 00007f616824e766 sp 00007ffe7adc38d0 error 5 in libtalloc.so.2.1.13[7f6168249000+e000]“

4. Downgrade of sssd not possible. Got yum errors that downgrade not possible. ( strange... )
5. yum reinstall libtalloc ( there was no update offered for newer revision )
5. Perform "yum update samba" and therefore all installed samba related packages
6. systemctl restart sssd + smb fixed login issues
Additional Information- Mixed sssd / samba version seem to produce libtalloc segfaults
- Should there be a warning popping up if only sssd will be updated that samba is needed as well ?


Actual installed Versions of SSSD:

rpm -qa |grep sssd
sssd-common-pac-1.16.2-13.el7_6.5.x86_64
sssd-krb5-1.16.2-13.el7_6.5.x86_64
sssd-libwbclient-1.16.2-13.el7_6.5.x86_64
sssd-krb5-common-1.16.2-13.el7_6.5.x86_64
sssd-ipa-1.16.2-13.el7_6.5.x86_64
sssd-ldap-1.16.2-13.el7_6.5.x86_64
sssd-proxy-1.16.2-13.el7_6.5.x86_64
sssd-1.16.2-13.el7_6.5.x86_64
sssd-client-1.16.2-13.el7_6.5.x86_64
sssd-common-1.16.2-13.el7_6.5.x86_64
sssd-ad-1.16.2-13.el7_6.5.x86_64
python-sssdconfig-1.16.2-13.el7_6.5.noarch


Actual installed Versions of LIBTALLOC:
rpm -qa |grep libtalloc
libtalloc-2.1.13-1.el7.x86_64


Actual installed Versions of SAMBA:

rpm -qa |grep samba
samba-common-4.8.3-4.el7.noarch
samba-libs-4.8.3-4.el7.x86_64
samba-common-libs-4.8.3-4.el7.x86_64
samba-common-tools-4.8.3-4.el7.x86_64
samba-client-libs-4.8.3-4.el7.x86_64
samba-4.8.3-4.el7.x86_64
TagsNo tags attached.
abrt_hash
URL

Activities

omerfsen

omerfsen

2019-05-01 11:05

reporter   ~0034411

I have exactly same issue. Upgrading package from

sssd-1.16.0-19.el7_5.8.x86_64 to
sssd-1.16.2-13.el7_6.5.x86_64 (with all dependant sub packages)

gives exceptions like:

1331.163526] sssd_be[2069]: segfault at ffffffffffffffa1 ip 00007f4456058766 sp 00007ffdd3879320 error 5 in libtalloc.so.2.1.13[7f4456053000+e000]
[ 2600.590953] sssd_be[2825]: segfault at ffffffffffffffa1 ip 00007f48c59c5766 sp 00007ffe6cce4690 error 5 in libtalloc.so.2.1.13[7f48c59c0000+e000]
[ 3313.988550] sssd_be[5072]: segfault at ffffffffffffffa1 ip 00007fd6ddd64d46 sp 00007fff994383a0 error 5 in libtalloc.so.2.1.10[7fd6ddd60000+f000]
mbosecke

mbosecke

2019-08-20 15:10

reporter   ~0035004

Are we able to increase the severity of this bug? The result a complete inability to authenticate with the system and seemingly no way of downgrading to resolve it.
TrevorH

TrevorH

2019-08-20 16:48

manager   ~0035006

It'll do you no good to increase the severity as CentOS only ships what RHEL ships and aims to be bug for bug compatible with it. To get this fixed you need RH to fix it in RHEL. If it hasn't been reported on bugzilla.redhat.com already then it needs to be.

Also, 7.5 is not current, 7.6 is. Are you up to date?

If you need to use yum downgrade then you have to specify ALL of the packages that need to be downgraded on the same command at the same time.

Issue History

Date Modified Username Field Change
2019-02-22 13:44 Henrik New Issue
2019-05-01 11:05 omerfsen Note Added: 0034411
2019-08-20 15:10 mbosecke Note Added: 0035004
2019-08-20 16:48 TrevorH Note Added: 0035006