"Gerald (Jerry) Carter" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > olly wrote: > | Hi Guys, > | > | I have experienced an issue when the smbd stops > | responding, and after cannot kill the process, using > | SIGTERM or SIGKILL. > > If you can't SIGKILL, it's a kernel bug IMO.
I kinda figured that the OS should be able to kill the process, but its the new samba that has exasibated the problem. I belive Jens' Suse 9.1 is also 2.4 kernel also. Is this correct? BTW i have noticed that samba as a seems better performannce since the last version i was using (3.0.6). Thanks! > > | A reboot cured the problem. I had the same problem > | once again a couple of days later. I realised that both > | times I had changed my smb.conf file directly, using > | vi, and shortly before the hang occured. The changes to > | the file were very minor, and the two were different > | options. Nothing appeared in the log, but my logging > | level is very low. Both times a reboot cured the problem. > > There changes to how we reload the printer cache between > 3.0.9 and 3.0.11. I've never seen the behavior you > describe but all my boxes are running 2.6 these days. Are you running SMP on any of these boxes, and I wonder if Jens is also? > > | > | Now here is my problem - the version of samba was > | tested before going live, but the problem was not > | picked ip, but now it is in a production environment, which > | I cannot disrupt until the last person goes home. I guess > | I what I am asking is what can you guys suggest I try > | in the small window i have to test it, to gather enough > | information for you to diagnose where the problem lies, if > | indeed it is related to the modified smd.conf. I will turn > | the logging level up to 10, but will have to stop samba > | first, so hopefully editing the smb.conf shouldn't cause > | a problem then. > > You can run smbcontrol smbd debug 10 to set the debug level. > You might also want to run strace on the 'hung' process > (although if it is really hung, you won't get any feedback > here either). Is changing the debug level at all likely to cause the same hanging as reloading the smb.conf or are the two operations sufficiently isolated that i can feel safe doing it with users online (at my own risk :-) ? > > | The server is a dual Zeon with Suse 9.0 kernel 2.4.21-273-smp4g > | with yast updates and samba 3.0.11. > | > | BTW, what is the actually policy for samba to watch for > | config updates? Is there a polling interval of sorts? > > internal checks every 3 minutes. grep for SMBD_RELOAD_CHECK. Thanks! > > > > > > > > cheers, jerry > ===================================================================== > Alleviating the pain of Windows(tm) ------- http://www.samba.org > GnuPG Key ----- http://www.plainjoe.org/gpg_public.asc > "I never saved anything for the swim back." Ethan Hawk in Gattaca > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.2.5 (GNU/Linux) > Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org > > iD8DBQFCGyo6IR7qMdg1EfYRAq8YAJ9R3YoMBiVchd5+9otFBfTSdnKYOgCfdVNt > ZChjVAhHuS7JGz8Vf5amX40= > =mId3 > -----END PGP SIGNATURE----- > -- > To unsubscribe from this list go to the following URL and read the > instructions: https://lists.samba.org/mailman/listinfo/samba > -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/listinfo/samba
