From: Wasim Bashir on 23 Dec 2009 07:00 Hi, I have replicated this on a test box, if you do a net cache flush, then restart samba and winbind, run getent passwd (only displays local users) then net cache list (will display all cache of remote users) The only way i know to fix this is to rename idmap config <name> and restart samba/winbind... but a week later the problem will be back.. seems strange to me, is this a bug with 3.3.9 or am i missing something here ? Thanks, Wasim 2009/12/22 Gaiseric Vandal <gaiseric.vandal(a)gmail.com> > I have similar issues with samba 3.0.37 on Solaris 10. I use winbind and > ldap for domain trusts (not for the users with in the domain.) Increasing > "idmap cache time" may reduce how often you need to reset things. When the > cache time expires I have to zap idmap entries from ldap and zap the idmap > cache tbd files. It appears samba can create the cache info but not > properly update or reread it once the cache has expired. > > I have been testing 3.4.3 and it seems better but I can't say for sure yet. > (Getting samba compiled with ldap and zfs support for Solaris is tricky.) > > > > > > > On 12/22/09 10:44, Wasim Bashir wrote: > >> Hi, >> >> I am having a weird issue with samba where once a week approximately at >> the >> same time users will lose connectivity, >> >> if i run >> >> wbinfo -u all users are displayed >> wbinfo -g all groups are displayed >> >> However running getent passwd only shows local-users, no remote users are >> shown.. >> >> To fix the issue I have to change the name of my idmap config and restart >> samba and winbind and everything works fine for a week... >> >> Am I missing something obvious here ? I have attached my config below : >> >> >> >> [global] >> security = ads >> max mux = 16384 >> log file = /home/sites/samba-log/log.%m >> >> ldap timeout = 45 >> ldap connection timeout = 30 >> max open files = 100000 >> realm = merlin.internaloffice.co.uk >> password server = 10.0.9.0 >> workgroup = WEBHOSTING >> idmap backend = tdb >> idmap uid = 500-2000000 >> idmap gid = 500-2000000 >> winbind enum users = yes >> winbind enum groups = yes >> template homedir = /home/sites/%U >> template shell = /bin/bash >> client use spnego = yes >> client ntlmv2 auth = yes >> encrypt passwords = yes >> winbind use default domain = yes >> winbind nss info = template rfc2307 >> restrict anonymous = 2 >> idmap config WEBHOSTING : schema_mode = rfc2307 >> idmap config WEBHOSTING : backend = ad >> idmap config WEBHOSTING : range = 500 - 300000000 >> >> >> >> >> [home] >> hide dot files = no >> path = /home/sites >> read only = no >> dos filetime resolution = yes >> >> I am using samba 3.3.9, do we know whether this issue has been fixed in >> samba 3.4.x ? >> >> Any help greatly appreciated. >> >> Thanks, >> >> Wasim >> >> > > -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
From: Moray Henderson on 23 Dec 2009 07:10 Wasim Bashir wrote: >I am having a weird issue with samba where once a week approximately at >the >same time users will lose connectivity, > >if i run > >wbinfo -u all users are displayed >wbinfo -g all groups are displayed > >However running getent passwd only shows local-users, no remote users are >shown.. > >To fix the issue I have to change the name of my idmap config and restart >samba and winbind and everything works fine for a week... > >Am I missing something obvious here ? I have attached my config below : > > > >[global] > security = ads > max mux = 16384 > log file = /home/sites/samba-log/log.%m > > ldap timeout = 45 > ldap connection timeout = 30 > max open files = 100000 > realm = merlin.internaloffice.co.uk > password server = 10.0.9.0 > workgroup = WEBHOSTING > idmap backend = tdb > idmap uid = 500-2000000 > idmap gid = 500-2000000 > winbind enum users = yes > winbind enum groups = yes > template homedir = /home/sites/%U > template shell = /bin/bash > client use spnego = yes > client ntlmv2 auth = yes > encrypt passwords = yes > winbind use default domain = yes > winbind nss info = template rfc2307 > restrict anonymous = 2 > idmap config WEBHOSTING : schema_mode = rfc2307 > idmap config WEBHOSTING : backend = ad > idmap config WEBHOSTING : range = 500 - 300000000 > > > > >[home] >hide dot files = no >path = /home/sites >read only = no >dos filetime resolution = yes > >I am using samba 3.3.9, do we know whether this issue has been fixed in >samba 3.4.x ? > >Any help greatly appreciated. > >Thanks, > >Wasim Could it be a network issue rather than Samba itself - a switch being turned off briefly, IP address being refreshed, DNS issue - that breaks the communication with kerberos or PDC? I heard of one site whose network was interrupted at the same time each day, which they eventually traced to a heavy delivery lorry crushing a badly-installed underground cable. Moray. "To err is human. To purr, feline" -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
|
Pages: 1 Prev: [Samba] net rpc user add produces duplicate SID Next: [Samba] Samba in win neighborhood Update |