[Samba] Odd high load occurance 2.2.4

Peter Polkinghorne pjmp at kingsfords.co.uk
Mon May 27 09:12:02 GMT 2002


I have Samba 2.2.4 installed for around 2 weeks - this seems to have fixed
the problems we had with 2.2.2 and runaway processes.

Today we had a Samba daemon runnig with a consistent 20% load (on a dual
processor SPARC with Solaris 2.6.  This machine acts as our PDC & WINS
server.

I used smbcontrol to up the debug levels from 1 to 2, then 3 - at 3
lots of:

[2002/05/27 16:54:12, 3, pid=15255, effective(117, 102), real(0, 0)] 
smbd/pipes.c:reply_pipe_write_and_X(198)
  writeX-IPC pnum=7547 nwritten=4280
[2002/05/27 16:54:12, 3, pid=15255, effective(117, 102), real(0, 0)] 
smbd/process.c:process_smb(866)
  Transaction 17100166 of length 4348
[2002/05/27 16:54:12, 3, pid=15255, effective(117, 102), real(0, 0)] 
smbd/process.c:switch_message(673)
  switch message SMBwriteX (pid 15255)
[2002/05/27 16:54:12, 3, pid=15255, effective(117, 102), real(0, 0)] 
smbd/sec_ctx.c:set_sec_ctx(314)
  setting sec ctx (117, 102) - sec_ctx_stack_ndx = 0
[2002/05/27 16:54:12, 3, pid=15255, effective(117, 102), real(0, 0)] 
smbd/sec_ctx.c:set_sec_ctx(319)
  5 user groups: 
  102 10 100 114 115 

sequences.  Any clues as to what this would be?

The client is a Windows 2000 server acting as a terminal server.
I have got some level 10 debug log.

-- 
Peter Polkinghorne, IT Manager		Kingsford Stacey Blackwell
ppolkinghorne at kingsfords.co.uk		14 Old Square, Lincoln's Inn
(44) 020 7447 1200			London WC2A 3UB

A list of names of the partners is open to inspection at the above address.






More information about the samba mailing list