Dear Members,
I would like to share my success story with a scheduled samba4 upgrade.
Our faculty domain contains one dc and two fileserver, plus the pc labor
windows machines(7 room * 25 machines). We use debian on the servers and
installed samba4 from stable backports.
Experiences:
* i could successfully apply the samdb.py and user.py pathces from
Rowland Penny. After that, the user creation worked as before
* for the first dc start i have to wait about 20 minutes. It was an
unpleasant suprise, i restarted samba many times, i thought i was stuck
into some loop.
At loglevel 10 i only saw some ldb operations and one cpu core went to
100% Example follows:
ldb: ldb_trace_next_request: (rdn_name)->modify
ldb: ldb_trace_next_request: (repl_meta_data)->modify
ldb: ldb_trace_next_request: (linked_attributes)->modify
ldb: ldb_trace_next_request: (partition)->modify
ldb: ldb_trace_next_request: (tdb)->modify
ldb: ldb_trace_next_request: (tdb)->search
ldb: ldb_trace_next_request: (schema_data)->del
ldb: ldb_trace_next_request: (tdb)->del
ldb: ldb_trace_next_request: (tdb)->del
ldb: ldb_trace_next_request: (tdb)->del
ldb: ldb_trace_next_request: (tdb)->del
ldb: ldb_trace_next_request: (tdb)->del
ldb: ldb_trace_next_request: (schema_data)->add
ldb: ldb_trace_next_request: (tdb)->add
ldb: ldb_trace_next_request: (tdb)->add
ldb: ldb_trace_next_request: (tdb)->add
I connected to the process with strace and i only saw this:
brk(0x7f0265025000) = 0x7f0265025000
brk(0x7f0265081000) = 0x7f0265081000
brk(0x7f02650dc000) = 0x7f02650dc000
brk(0x7f0265138000) = 0x7f0265138000
I think it was some kind of ldb upgrade/maintenance. It would be handy
to have some feedback or progess bar. I really don't like unresponsive
starts, especially after an upgrade...
If needed, i could past the whole debug log/samba config/etc...
* apart from the previous problem, the 4.1.17 version works as expected!
Regards,
J?nos