File Server access nightmare
in [UK Linux]
|
Prev: biew-6.1.0 has been released
Next: Installation Proglem
From: Alan Secker on 15 Dec 2009 05:31 For several days I have been experiencing difficulties in connecting to my firm's file-server. This is a six work-station LAN using samba to access common files. Shortly after replacing my work-station with a newer machine the problems began. Today I decided to cold-boot the file server and log the results. This is what happened: 1) Rebooted server. 2) Powered up work-stations. None could 'see' the server. 3) Executed samba restart on the server 4) All users could now ping the server 5) Everyone except me could now 'see' and access common files on the server via smb4k. 6) As I have had to do every morning for the last four working days, on the file server, I ran smbpasswd and re-entered my samba username and password. 7) Checked again via smb4k and could now access the server's files. I've wracked my brains as to what may be causing the problem but my knowledge is peripheral and my experience, scant. I would greatly appreciate some experienced prodding.
From: Dave Farrance on 15 Dec 2009 07:06 Alan Secker <alan(a)asandco.co.uk> wrote: >For several days I have been experiencing difficulties in connecting to my >firm's file-server. This is a six work-station LAN using samba to access >common files. Check the smb and nmb log files on the server. They should tell you why a given connection is refused. You can set various levels of verbosity for the logging via the config file. -- Dave Farrance
From: Martin Gregorie on 15 Dec 2009 08:23 On Tue, 15 Dec 2009 10:31:33 +0000, Alan Secker wrote: > 6) As I have had to do every morning for the last four working days, on > the > file server, I ran smbpasswd and re-entered my samba username and > password. > 7) Checked again via smb4k and could now access the server's files. > Some things to check, in approximate order of decreasing likeliness: - Does testparm show any errors in smb.conf? - Does restarting Samba cause the same problem as rebooting the server? - Is it a file permissions issue? If so, it may show up in the system log. - Could this be an SElinux/appArmor issue? If it is installed, does disabling it fix the problem? - Is the SAMBA password file in a sensible place, i.e. not on a ramdisk or in /tmp? -- martin@ | Martin Gregorie gregorie. | Essex, UK org |
From: Alan Secker on 15 Dec 2009 10:49 Well, the log files didn't reveal anything suspect. testparm was happy. Restarting samba did not cause the same problems as rebooting. File permissions were an unlikely suspect as everyone but me had complete access. I checked my $HOME anyway.It was OK. Neither SELiniux or appArmor had been deployed smb.conf is at /etc/samba BUT, the server's clock was an hour out! I changed it. The only other change I made was to the verbosity level for the logs. I raised it by 1. After lunch today I had a window of one hour to be able to close down the whole network and run it up bit by bit. Lo' and behold, it all ran smoothly. I did it twice more to be certain and it seems as solid as a week ago. With no real explanation, I still feel exposed but thanks for your advice. Sometimes knowing someone is at your shoulder can feel very supportive.
From: Geoffrey Clements on 15 Dec 2009 12:01
"Alan Secker" <alan(a)asandco.co.uk> wrote in message news:ZdudnX4NO-CgLbrWnZ2dnUVZ8kadnZ2d(a)pipex.net... > Well, the log files didn't reveal anything suspect. > testparm was happy. > Restarting samba did not cause the same problems as rebooting. > File permissions were an unlikely suspect as everyone but me had complete > access. I checked my $HOME anyway.It was OK. > Neither SELiniux or appArmor had been deployed > smb.conf is at /etc/samba > > BUT, the server's clock was an hour out! One hour - that seems suspicious. Some questions come to mind: Is the server's clock on UTC or localtime? What time zone is the server set to? It's worth checking this or you might have the same problem again (in March maybe). -- Geoff |