NFS setup
Hi all, I´ve read other threads on this and posted some of my problems. As it stands i am stuck, so forgive me for starting a new thread in what i think is the right section. My aim is to share my backup folder to the other computers on my home network. The location of this folder is home/backup on the server machine. I have installed nfs to the server via.. sudo apt-get install nfs-common portmap nfs-kernel-server I have installed nfs to the client via.. sudo apt-get install nfs-common portmap My desktop is the server machine 192.168.1.2 My laptop is the client machine 192.168.1.3 I have edited etc/exports to share /home/backup /home/backup 192.168.1.3(rw,sync,no_root_squash) I would like to know if i have done everything properly on the server and client machine. Because i got the error bellow when i try to mount the nfs partition on the client side. $ sudo mount 192.168.1.2:/home/backup /mnt/nfs $ mount.nfs: 192.168.1.2:/home/backup failed, reason given by server: Permission denied i can ping between each machine. I believe the permission is denied because both machines don´t have each other listed at /etc/hosts /etc/hosts.allow /etc/hosts.deny Is that possible? and how to add them properly? i tried to add them but it doesn´t work either. please help, iuri -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup
On Wed, 12 Dec 2007 17:00:20 +0100, iuri de araujo sampaio wrote: I have installed nfs to the server via.. sudo apt-get install nfs-common portmap nfs-kernel-server I have installed nfs to the client via.. sudo apt-get install nfs-common portmap My desktop is the server machine 192.168.1.2 My laptop is the client machine 192.168.1.3 I have edited etc/exports to share /home/backup /home/backup 192.168.1.3(rw,sync,no_root_squash) Looking good so far. Did you restart nfs-kernel-server after editing /etc/exports ? Have you checked that portmap mountd are running on the server the client? Does /var/log/syslog on the server show anything interesting? ... and now for the random one ... Do you run bind on the same machine as your nfs server? This has a port conflict somewhere with portmap, so ensure portmap starts before bind. -- Stephen Patterson :: [EMAIL PROTECTED] :: http://patter.mine.nu/ GPG: B416F0DE :: Jabber: [EMAIL PROTECTED] Don't be silly, Minnie. Who'd be walking round these cliffs with a gas oven? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup
Hi, On Wed, Dec 12, 2007 at 12:56:57PM -0300, iuri de araujo sampaio wrote: I would like to know if i have done everything properly on the server and client machine. did you do exportfs -a or restart/reload nfs-kernel-server after editing /etc/exports? i can ping between each machine. I believe the permission is denied because both machines don´t have each other listed at /etc/hosts /etc/hosts.allow /etc/hosts.deny Is that possible? and how to add them properly? i tried to add them but it doesn´t work either. yes, that is possible. if server and client can't resolv each others names mounting will be painfully slow, because nfsd tries to resolv the hostname. setting up an dnsd or editing /etc/hosts is not requiered but you should do it anyways. if you deny connects in /etc/hosts.deny you need to put something like this in /etc/hosts.allow logd,mountd,statd,portmap: 192.168.1.0/24 have a look at iptables -nv -L too, if it still doesn't work after that. greetings Peter Werner -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup
Do user names, uid, and gid match on the NFS server and the client? This will also have to be true for group names and uid. On Wednesday 12 December 2007 14:10:09 Peter Werner wrote: Hi, On Wed, Dec 12, 2007 at 12:56:57PM -0300, iuri de araujo sampaio wrote: I would like to know if i have done everything properly on the server and client machine. did you do exportfs -a or restart/reload nfs-kernel-server after editing /etc/exports? i can ping between each machine. I believe the permission is denied because both machines don´t have each other listed at /etc/hosts /etc/hosts.allow /etc/hosts.deny Is that possible? and how to add them properly? i tried to add them but it doesn´t work either. yes, that is possible. if server and client can't resolv each others names mounting will be painfully slow, because nfsd tries to resolv the hostname. setting up an dnsd or editing /etc/hosts is not requiered but you should do it anyways. if you deny connects in /etc/hosts.deny you need to put something like this in /etc/hosts.allow logd,mountd,statd,portmap: 192.168.1.0/24 have a look at iptables -nv -L too, if it still doesn't work after that. greetings Peter Werner
RE: NFS Setup/Configuration (WORKING)
SPIDAS ! YOU DA'MAN !!! Thank you SO much...I followed your steps and they worked PERFECTLY !!! Not to mention...you little tutorial here is better than anything I looked at on the net. You hit every piece DEAD-ON. I can not thank you enough... Thanks for everything I will be saving this message. -Stephen From: s spidas [mailto:[EMAIL PROTECTED] Sent: Friday, November 17, 2006 11:05 PM To: Stephen Yorke Cc: Debian User Subject: Re: NFS Setup/Configuration It's quite straight forward actually 1. Make sure 'nfs-common' is installed on all the clients (I'm assuming all clients will be linux boxes),and that both 'nfs-common' and 'nfs-kernel-server' are installed on all servers. (apt-get install file will solve dependencies if any). 2. Modify the following files (3 of 'em) on the server; A: /etc/exports; B: /etc/hosts.allow C: /etc/hosts.deny A. /etc/exports (this is the access control list for all filesystems to be exported) e.g. to export /home as read/write to computers on 10.10.0.0 /home 10.10.0.0/255.255.255.0(rw,no_root_squash,no_subtree_check) (will give access to /home to all computers on network 10.10.0.0) (note that there must be no white space between the options!) (for explanations see man nfs) B. /etc/hosts.allow (list of hosts that are allowed to access the system). To protect your system, define specific service access. Avoid giving blanket access to all. You may use the following: portmap: 10.10.0.0/255.255.255.0 lockd: 10.10.0.0/255.255.255.0 mountd: 10.10.0.0/255.255.255.0 rquotad: 10.10.0.0/255.255.255.0 statd: 10.10.0.0/255.255.255.0 sshd: 10.10.0.0/255.255.255.0 C. /etc/hosts.deny (list of hosts that are NOT allowed to access the system. ALL: ALL 3. Now force a re-read of the nfs/exportfs by issuing the following command as root: #: exportfs -ra 4. You may check that things are working ok : rpcinfo -p 5. Go to a client, and as #: mount server:/share mount point. e.g. #: mount 10.10.0.200:/home /mnt GOODLUCK! On 11/16/06, Stephen Yorke [EMAIL PROTECTED] wrote: All, I have been trying to setup NFS on my home Debian systems. I have followed all the stuff from nfs.sourceforge.net but I keep getting Permission Denied errors when I try to mount the Exported folder. Here is what I have... :: PXE Server :: tftpd-hpa pxe nfs-user-server I know that my TFTP/PXE services are working properly because I network boot a PC and I get the boot prompt/menu. :: exports :: /tftpboot/debian-live 10.1.1.0/255.255.255.240(ro,no_root_squash) http://10.1.1.0/255.255.255.240%28ro,no_root_squash%29 :: hosts.allow :: ALL:10.1.1.0/27 :: hosts.deny :: 'empty' When I try this... mount -t nfs 10.1.1.23:/tftpboot/debian-live /nfstest I get the error... mount: 10.1.1.23:/tftpboot/debian-live failed, reason given by server: Permission denied Can anyone shed any light on this subject? If I run tcpdump on the PXE/TFTP/NFS server I see the client trying to connect but it never gets connected. I do not see anything in any of my log files which would help me out either... Thanks in advance, Stephen
Re: NFS Setup/Configuration
Jesus Arocho wrote: 0.02: I am running an NFS server on a debian box; 5 kubuntu workstations mount NFS exports from the box. The passwords, login names, login ID number, group names and group ID numbers must match between the workstation and the server. My two cents: Booting from NFS and just mounting ordinary NFS shares once the system is already running are two completely different things with nothing in common but the name 'NFS'. -- Address:Daniel Baumann, Burgunderstrasse 3, CH-4562 Biberist Email: [EMAIL PROTECTED] Internet: http://people.panthera-systems.net/~daniel-baumann/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: NFS Setup/Configuration
I have not thought of this but I am logged in with root from my workstation which does in fact have the same password as the one on the NFS server. Also, I checked my PORTMAP file and it is not listening on the loopback address (127.0.0.1). That line is in fact commented out. -Stephen -Original Message- From: Jesus Arocho [mailto:[EMAIL PROTECTED] Sent: Thursday, November 16, 2006 5:59 PM To: debian-user@lists.debian.org Subject: Re: NFS Setup/Configuration On Thursday 16 November 2006 12:36, Stephen Yorke wrote: All, I have been trying to setup NFS on my home Debian systems. I have followed all the stuff from nfs.sourceforge.net but I keep getting Permission Denied errors when I try to mount the Exported folder. Here is what I have... :: PXE Server :: tftpd-hpa pxe nfs-user-server I know that my TFTP/PXE services are working properly because I network boot a PC and I get the boot prompt/menu. :: exports :: /tftpboot/debian-live 10.1.1.0/255.255.255.240(ro,no_root_squash) :: hosts.allow :: ALL:10.1.1.0/27 :: hosts.deny :: 'empty' When I try this... mount -t nfs 10.1.1.23:/tftpboot/debian-live /nfstest I get the error... mount: 10.1.1.23:/tftpboot/debian-live failed, reason given by server: Permission denied Can anyone shed any light on this subject? If I run tcpdump on the PXE/TFTP/NFS server I see the client trying to connect but it never gets connected. I do not see anything in any of my log files which would help me out either... Thanks in advance, Stephen 0.02: I am running an NFS server on a debian box; 5 kubuntu workstations mount NFS exports from the box. The passwords, login names, login ID number, group names and group ID numbers must match between the workstation and the server. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS Setup/Configuration
It's quite straight forward actually 1. Make sure 'nfs-common' is installed on all the clients (I'm assuming all clients will be linux boxes),and that both 'nfs-common' and 'nfs-kernel-server' are installed on all servers. (apt-get install file will solve dependencies if any). 2. Modify the following files (3 of 'em) on the server; A: /etc/exports; B: /etc/hosts.allow C: /etc/hosts.deny A. /etc/exports (this is the access control list for all filesystems to be exported) e.g. to export /home as read/write to computers on 10.10.0.0 /home 10.10.0.0/255.255.255.0(rw,no_root_squash,no_subtree_check) (will give access to /home to all computers on network 10.10.0.0) (note that there must be no white space between the options!) (for explanations see man nfs) B. /etc/hosts.allow (list of hosts that are allowed to access the system). To protect your system, define specific service access. Avoid giving blanket access to all. You may use the following: portmap: 10.10.0.0/255.255.255.0 lockd: 10.10.0.0/255.255.255.0 mountd: 10.10.0.0/255.255.255.0 rquotad: 10.10.0.0/255.255.255.0 statd: 10.10.0.0/255.255.255.0 sshd: 10.10.0.0/255.255.255.0 C. /etc/hosts.deny (list of hosts that are NOT allowed to access the system. ALL: ALL 3. Now force a re-read of the nfs/exportfs by issuing the following command as root: #: exportfs -ra 4. You may check that things are working ok : rpcinfo -p 5. Go to a client, and as #: mount server:/share mount point. e.g. #: mount 10.10.0.200:/home /mnt GOODLUCK! On 11/16/06, Stephen Yorke [EMAIL PROTECTED] wrote: All, I have been trying to setup NFS on my home Debian systems. I have followed all the stuff from nfs.sourceforge.net but I keep getting Permission Denied errors when I try to mount the Exported folder. Here is what I have… :: PXE Server :: tftpd-hpa pxe nfs-user-server I know that my TFTP/PXE services are working properly because I network boot a PC and I get the boot prompt/menu. :: exports :: /tftpboot/debian-live 10.1.1.0/255.255.255.240(ro,no_root_squash)http://10.1.1.0/255.255.255.240%28ro,no_root_squash%29 :: hosts.allow :: ALL:10.1.1.0/27 :: hosts.deny :: 'empty' When I try this… mount -t nfs 10.1.1.23:/tftpboot/debian-live /nfstest I get the error… mount: 10.1.1.23:/tftpboot/debian-live failed, reason given by server: Permission denied Can anyone shed any light on this subject? If I run tcpdump on the PXE/TFTP/NFS server I see the client trying to connect but it never gets connected. I do not see anything in any of my log files which would help me out either... Thanks in advance, Stephen
Re: NFS Setup/Configuration
I have had an ongoing issue with the later versions of mount which was claimed to have been fixed in bug 354075: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=354075 I have held mount at version 2.12r-6 as anything later refuses to work with my NFS server, also running Sid. The latest version I've tried is 2.12r-12 on 25 Oct and I even tried with the (then) latest kernel, 2.6.18 on the laptop (the server was at 2.6.18) and found no change. I would like to hear from anyone who is experiencing the same problem. As an aside, I am running Shorewall on the server and Firehol on the laptop to configure Netfilter. Still, disabling both firewalls has no effect for me on the failure. Surely, I can't be the only one in the Debian Universe with this problem. ;-) - Nate -- Wireless | Amateur Radio Station N0NB | Successfully Microsoft Amateur radio exams; ham radio; Linux info @ | free since January 1998. http://www.qsl.net/n0nb/ | Debian, the choice of My Kawasaki KZ-650 SR @| a GNU generation! http://www.networksplus.net/n0nb/ | http://www.debian.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
NFS Setup/Configuration
All, I have been trying to setup NFS on my home Debian systems. I have followed all the stuff from nfs.sourceforge.net but I keep getting Permission Denied errors when I try to mount the Exported folder. Here is what I have... :: PXE Server :: tftpd-hpa pxe nfs-user-server I know that my TFTP/PXE services are working properly because I network boot a PC and I get the boot prompt/menu. :: exports :: /tftpboot/debian-live 10.1.1.0/255.255.255.240(ro,no_root_squash) :: hosts.allow :: ALL:10.1.1.0/27 :: hosts.deny :: 'empty' When I try this... mount -t nfs 10.1.1.23:/tftpboot/debian-live /nfstest I get the error... mount: 10.1.1.23:/tftpboot/debian-live failed, reason given by server: Permission denied Can anyone shed any light on this subject? If I run tcpdump on the PXE/TFTP/NFS server I see the client trying to connect but it never gets connected. I do not see anything in any of my log files which would help me out either... Thanks in advance, Stephen
Re: NFS Setup/Configuration
On Thursday 16 November 2006 12:36, Stephen Yorke wrote: All, I have been trying to setup NFS on my home Debian systems. I have followed all the stuff from nfs.sourceforge.net but I keep getting Permission Denied errors when I try to mount the Exported folder. Here is what I have... :: PXE Server :: tftpd-hpa pxe nfs-user-server I know that my TFTP/PXE services are working properly because I network boot a PC and I get the boot prompt/menu. :: exports :: /tftpboot/debian-live 10.1.1.0/255.255.255.240(ro,no_root_squash) :: hosts.allow :: ALL:10.1.1.0/27 :: hosts.deny :: 'empty' When I try this... mount -t nfs 10.1.1.23:/tftpboot/debian-live /nfstest I get the error... mount: 10.1.1.23:/tftpboot/debian-live failed, reason given by server: Permission denied Can anyone shed any light on this subject? If I run tcpdump on the PXE/TFTP/NFS server I see the client trying to connect but it never gets connected. I do not see anything in any of my log files which would help me out either... Thanks in advance, Stephen 0.02: I am running an NFS server on a debian box; 5 kubuntu workstations mount NFS exports from the box. The passwords, login names, login ID number, group names and group ID numbers must match between the workstation and the server. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
nfs setup on woody
I've recently moved a third machine to Debian. This one is Woody, base system installed from first disk of CD set and then from ftp. Only problem is I cannot connect to it by nfs. On the woody system, I get this when nfs starts: Exporting directories for NFS kernel daemon...machine.domain:/home\ /richard:Function not implemented done. Starting NFS kernel daemon: nfsd mountd. # rpcinfo -p program vers proto port 102 tcp111 portmapper 102 udp111 portmapper 3910021 tcp 1025 sgi_fam 3910022 tcp 1025 sgi_fam 1000241 udp 1037 status 1000241 tcp 2006 status 151 udp 1038 mountd 151 tcp 2007 mountd 152 udp 1038 mountd 152 tcp 2007 mountd Which is odd, as I expected 'nfs' and 'nlockmgr' instead of 'sgi_fam'. What is sgi_fam, anyway? And what package installed it? Sounds like Silicon Graphics, in which case it's presumably a mistake. In view of the Function not implemented message, it's hardly a surprise that I can't connect. This is what I get when I try the mount on my machine (running Sid FWIW): $ mount attic mount: machine.domain:/home/richard failed, reason given by server:\ Permission denied I've drawn a blank in the howto and removing and reinstalling nfs makes no difference. Any wise words gratefully accepted... -- richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Sat, Jan 03, 2004 at 11:49:36AM +, Richard Lyons [EMAIL PROTECTED] wrote: On the woody system, I get this when nfs starts: Exporting directories for NFS kernel daemon...machine.domain:/home\ /richard:Function not implemented Can you show us the exact line for this share from the exports file? /GCS -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Saturday 03 January 2004 14:22, GCS wrote: On Sat, Jan 03, 2004 at 11:49:36AM +, Richard Lyons [EMAIL PROTECTED] wrote: On the woody system, I get this when nfs starts: Exporting directories for NFS kernel daemon...machine.domain:/home\ /richard:Function not implemented Can you show us the exact line for this share from the exports file? /GCS yes: /home/richard/ 192.168.0.2(rw,sync) very much like some others that worked fine. -- richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Sat, Jan 03, 2004 at 03:22:09PM +, Richard Lyons [EMAIL PROTECTED] wrote: yes: /home/richard/ 192.168.0.2(rw,sync) Thanks. Well, I haven't seen the 'sync' option yet, and it's not mentioned in the man page either - but I am not an NFS guru, it can be good. Anyway, it may worth a try to remove it, even if my Sarge box re-started the nfsd without any warnings with this option. very much like some others that worked fine. Do you have more shares on this machine, or is this the only one in the exports file? Cheers, GCS -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Saturday 03 January 2004 15:34, GCS wrote: On Sat, Jan 03, 2004 at 03:22:09PM +, Richard Lyons [EMAIL PROTECTED] wrote: yes: /home/richard/ 192.168.0.2(rw,sync) Thanks. Well, I haven't seen the 'sync' option yet, and it's not mentioned in the man page either - but I am not an NFS guru, it can be good. Anyway, it may worth a try to remove it, even if my Sarge box re-started the nfsd without any warnings with this option. I have tried it with and without 'sync' (I have no recollection where I copied it from originally). Also with machine.domain in place of IP. Also tried (ro) in place of (rw). Same error message each time: # exportfs -ra machine.domain:/home/richard: Function not implemented Do you have more shares on this machine, or is this the only one in the exports file? It is the only one at the moment. I used the same pattern with various exports on all three machines previously. The Woody machine was then running RH8 and the other two are Sid. All used to respond to # rpcinfo -p with something like program vers proto port 102 tcp111 portmapper 102 udp111 portmapper 132 udp 2049 nfs 133 udp 2049 nfs 132 tcp 2049 nfs 133 tcp 2049 nfs 1000211 udp 32770 nlockmgr 1000213 udp 32770 nlockmgr 1000214 udp 32770 nlockmgr 1000211 tcp 32768 nlockmgr 1000213 tcp 32768 nlockmgr 1000214 tcp 32768 nlockmgr 151 udp897 mountd 151 tcp900 mountd 152 udp897 mountd 152 tcp900 mountd 153 udp897 mountd 153 tcp900 mountd 1000241 udp 1009 status 1000241 tcp 1012 status which is why I expected some 'nfs' and 'nlockmgr' lines rather than those 'sgi_fam' lines. I can't help thinking this is a problem of the installation rather than the /etc/exports file. I installed nfs-common and nfs-kernel-server but perhaps I am missing something else? Or are there other config files that could be misconfigured? hosts.allow has my IP listed for portmap, lockd, mountd, rquotad, statd, and rpc. In any case that 'Function not implemented' suggests something else (I don't know what!) . -- richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Sat, Jan 03, 2004 at 04:40:10PM +, Richard Lyons [EMAIL PROTECTED] wrote: Do you have more shares on this machine, or is this the only one in the exports file? It is the only one at the moment. Then I am almost sure that the kernel support is missing. Do you have the following modules loaded: nfs, lockd, sunrpc? It's also ok to have them compiled into the kernel: CONFIG_NFS_FS, CONFIG_NFSD, CONFIG_LOCKD, CONFIG_SUNRPC. BTW, what's your kernel version? which is why I expected some 'nfs' and 'nlockmgr' lines rather than those 'sgi_fam' lines. FAM stands for File Alternation Monitor, developed by SGI Inc (previously for their desktop applications since ~Irix 6.3). I can't help thinking this is a problem of the installation rather than the /etc/exports file. Check the kernel support mentioned above. hosts.allow has my IP listed for portmap, lockd, mountd, rquotad, statd, and rpc. It does not matter - yet! - as it needs for connecting to your machine, but you have problems right with starting NFS services. Cheers, GCS -- BorsodChem Joint-Stock Company Linux Support Center Software engineer Developer +36-48-511211/12-99 +36-20-4441745 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Saturday 03 January 2004 16:42, GCS wrote: On Sat, Jan 03, 2004 at 04:40:10PM +, Richard Lyons [EMAIL PROTECTED] wrote: Do you have more shares on this machine, or is this the only one in the exports file? It is the only one at the moment. Then I am almost sure that the kernel support is missing. Do you have the following modules loaded: nfs, lockd, sunrpc? It's also ok to have them compiled into the kernel: CONFIG_NFS_FS, CONFIG_NFSD, CONFIG_LOCKD, CONFIG_SUNRPC. BTW, what's your kernel version? # uname -r 2.2.20-idepci no they are not loaded: # lsmod Module Size Used by parport_pc 7276 1 (autoclean) lp 4580 0 parport 6676 1 [parport_pc lp] af_packet 6136 0 (unused) and # modprobe nfs modprobe: Can't locate module nfs # modprobe lockd modprobe: Can't locate module lockd So you are evidently right. Pity. The howto suggests anything over 2.2.16 supports nfs. I did recently apt-get kernel-image 2.4.18-1-k6 (actually via aptitude), but it doesn't boot correctly, so I continue with the 2.2.20. I suppose I could try dist-upgrade. I know that nfs works out of the box in Sid, and I haven't had any real problems with Sid. The reason I chose Woody was that I just wanted to play safe as that box is used by the rest of the family, who don't want to cope with problems, particularly when I am out of the country. Someone will probably tell me to compile a kernel, but I've wasted days reading howtos and going through configurations I understand about 10% of, only to get unusable kernels, so I'm not inclined to go that route. -- richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Sat, Jan 03, 2004 at 05:34:46PM +, Richard Lyons [EMAIL PROTECTED] wrote: [snip] So you are evidently right. Pity. The howto suggests anything over 2.2.16 supports nfs. Yup, the kernel does support it, _if_ you choose to include it into the compilation. I think it just missed in that kernel. I did recently apt-get kernel-image 2.4.18-1-k6 (actually via aptitude), but it doesn't boot correctly, so I continue with the 2.2.20. What's was the problem? Have you set it up correctly (I mean initrd-tools is not enough, you have to specify the initrd.img in your bootloader configuration - probably in lilo.conf). I suppose I could try dist-upgrade. I know that nfs works out of the box in Sid, Just to be clear: dist-upgrade would give you Sarge on a Woody box, which is under testing. Someone will probably tell me to compile a kernel, but I've wasted days reading howtos and going through configurations I understand about 10% of, only to get unusable kernels, so I'm not inclined to go that route. It's not that hard, but you are right, it would be easier to install an image from deb than compile your one. Cheers, GCS -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: nfs setup on woody
On Saturday 03 January 2004 17:30, GCS wrote: On Sat, Jan 03, 2004 at 05:34:46PM +, Richard Lyons [EMAIL PROTECTED] wrote: [snip] I did recently apt-get kernel-image 2.4.18-1-k6 (actually via aptitude), but it doesn't boot correctly, so I continue with the 2.2.20. What's was the problem? Have you set it up correctly (I mean initrd-tools is not enough, you have to specify the initrd.img in your bootloader configuration - probably in lilo.conf). The relevent section of lilo.conf: default=Linux_2.2 image=/vmlinuz label=NoGo_2.4 read-only initrd=/initrd.img image=/vmlinuz_2.2 label=Linux_2.2 read-only optional and the simlinks are in place: initrd.img - /boot/initrd.img-2.4.18-1-k6 vmlinuz - boot/vmlinuz-2.4.18-1-k6 and lilo was happy. What was the problem? Well, hundreds of error messages, which are not in dmesg, inluding stacks of insmod warnings and IRQ conflicts, but it all goes too fast to read or remember. The system does start, and even KDE loads normally, but there is no networking at all. I didn't bother to investigate in detail because so many errors were listed that it seemed like a life's work. dmesg does include: PCI: No IRQ known for interrupt pin A of device 00:11.1. Please try using pci=biosirq. ... VP_IDE: Unknown VIA SouthBridge, contact Vojtech Pavlik [EMAIL PROTECTED] ... ide0 at 0x1f0-0x1f7,0x3f6 on irq 14 ide0: probed IRQ 14 failed, using default. ide1 at 0x170-0x177,0x376 on irq 15 ide1: probed IRQ 15 failed, using default. ... PCI: Found IRQ 3 for device 00:10.0 IRQ routing conflict for 00:10.0, have irq 9, want irq 3 IRQ routing conflict for 00:10.1, have irq 9, want irq 3 IRQ routing conflict for 00:10.2, have irq 9, want irq 3 IRQ routing conflict for 00:10.3, have irq 9, want irq 3 ...(and the same again for 00:10.1 and 00:10.2) but the insmod and other things are not in there. I suppose I could try dist-upgrade. I know that nfs works out of the box in Sid, Just to be clear: dist-upgrade would give you Sarge on a Woody box, which is under testing. What do you mean by on a Woody box - does the kernel not get upgraded? And if it would only take me to Sarge, I could do it twice to get Sid, couldn't I? Someone will probably tell me to compile a kernel, but I've wasted days reading howtos and going through configurations I understand about 10% of, only to get unusable kernels, so I'm not inclined to go that route. It's not that hard, but you are right, it would be easier to install an image from deb than compile your one. You probably understand the configuration options, which must help a little. Anyway, I've never succeeded yet. It doesn't mean I will never try again... Thanks for the help so far. -- richard -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: need help with NFS setup
* Paul E. Condon [EMAIL PROTECTED]:~$ mount /cmn mount: RPC: Port mapper failure - RPC: Unable to receive What does this mean? And how do I fix it? First, look into the nfs howto, you'll find it by searching for 'nfs' and 'howto' on the web. Then here is a few things to consider: Do you have firewall (iptables, ipchain, etc) between the computers? Is portmap running? (Try: 'ps aux | grep portmap' on both computers.) Try to mount an nfs partition locally. -- Jon Haugsand, [EMAIL PROTECTED] http://www.norges-bank.no -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: need help with NFS setup
On Mon, June 23 at 5:11 PM EDT Paul E Condon [EMAIL PROTECTED] wrote: I copy exports, hosts.allow, and hosts.deny from the old server to the new, and restart the daemons. Then I attempt to mount newserver:/home from the working client. I get an error message: [EMAIL PROTECTED]:~$ mount /cmn mount: RPC: Port mapper failure - RPC: Unable to receive What does this mean? And how do I fix it? There may be different packages needed if you aren't used to woody. I am on unstable and here are the ones I used: ii nfs-common 1.0.3-1NFS support files common to client and serve ii nfs-user-serve 2.2beta47-15 User space NFS server. Maybe you installed nfs-kernel-server instead and the kernel isn't configured for it? I used nfs-user-server and setup was trivial as you describe above. I'll assume you can ping each other, by hostname? Shawn Lamson [EMAIL PROTECTED] Debian GNU/Linux 3.0 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: need help with NFS setup SOLVED
Thanks to all. It turned out that I had done all the nfs stuff correctly, but I had selected the wrong driver for my rtl8139 card. I learned this when I thought to try to ping the new server from my test box. I should write myself an Avoid Dumb Mistakes HOWTO. The Debian install program said that the driver I chose installed OK. Turns out rtl8139-scyld doesn't really work for me, but 8139too does. On Tue, Jun 24, 2003 at 07:10:10AM -0400, Shawn Lamson wrote: On Mon, June 23 at 5:11 PM EDT Paul E Condon [EMAIL PROTECTED] wrote: I copy exports, hosts.allow, and hosts.deny from the old server to the new, and restart the daemons. Then I attempt to mount newserver:/home from the working client. I get an error message: [EMAIL PROTECTED]:~$ mount /cmn mount: RPC: Port mapper failure - RPC: Unable to receive What does this mean? And how do I fix it? There may be different packages needed if you aren't used to woody. I am on unstable and here are the ones I used: ii nfs-common 1.0.3-1NFS support files common to client and serve ii nfs-user-serve 2.2beta47-15 User space NFS server. Maybe you installed nfs-kernel-server instead and the kernel isn't configured for it? I used nfs-user-server and setup was trivial as you describe above. I'll assume you can ping each other, by hostname? Shawn Lamson [EMAIL PROTECTED] Debian GNU/Linux 3.0 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- Paul E Condon [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
need help with NFS setup
I am attempting to setup a replacement for a server that I have had running for a couple of years. Eventually, the replacement will have wonderful new features, but for now I am haveing trouble merely getting a minimal NFS working. The old system was potato, the new is woody. I have a client host that is woody and works with the old potato server. All software is current debian packages. I copy exports, hosts.allow, and hosts.deny from the old server to the new, and restart the daemons. Then I attempt to mount newserver:/home from the working client. I get an error message: [EMAIL PROTECTED]:~$ mount /cmn mount: RPC: Port mapper failure - RPC: Unable to receive What does this mean? And how do I fix it? TIA -- Paul E Condon [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: need help with NFS setup
El Martes, 24 de Junio de 2003 01:11, Paul E Condon escribió: I am attempting to setup a replacement for a server that I have had running for a couple of years. Eventually, the replacement will have wonderful new features, but for now I am haveing trouble merely getting a minimal NFS working. The old system was potato, the new is woody. I have a client host that is woody and works with the old potato server. All software is current debian packages. I copy exports, hosts.allow, and hosts.deny from the old server to the new, and restart the daemons. Then I attempt to mount newserver:/home from the working client. I get an error message: [EMAIL PROTECTED]:~$ mount /cmn mount: RPC: Port mapper failure - RPC: Unable to receive What does this mean? And how do I fix it? Have you setup the portmap service? Does it have access permission? Luis Llana. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
nfs setup question
I've setup nfs a couple of times with potato, but to check if nfs is running, I would issue a rpcinfo -p and usually get 2 or 3 lines from protmapper and status.d (or something like that) if nfs is not running, and if nfs is running the rpcinfo -p commandf would return about 15 lines including nfs and mountd. rpcinf -p servername now returns rpcinfo: can't contact portmapper RPC: remote system error - connection refused. what did I do wrong? what did I overlook? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
NFS setup problem
when I check to status of nfs on the server I';m trying to setup, the rpcinfo -p servername command returns rpcinfo: can't contact portmapper RPC: rompte system error - connection refused the lsmod command shows nfsd, nfs and lockd. what did I do wrong? what am I overlooking. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup problem
On 11 Feb 2003 08:56:06 -0700 Andreas J. Guelzow [EMAIL PROTECTED] wrote: On Tue, 2003-02-11 at 07:51, Phil Beder wrote: when I check to status of nfs on the server I';m trying to setup, the rpcinfo -p servername command returns rpcinfo: can't contact portmapper RPC: rompte system error - connection refused the lsmod command shows nfsd, nfs and lockd. what did I do wrong? what am I overlooking. Hi, did you remember to set up /etc/hosts.allow ? You should have something like: portmap:10.10.0.0/255.255.0.0 127.0.0.0/255.255.255.0 mountd:127.0.0.0/255.255.255.0 10.10.0.0/255.255.0.0 EXCEPT 10.10.10.10 statd:10.10.0.0/255.255.0.0 127.0.0.0/255.255.255.0 Andreas ~~~ thanks but I don't seem to have man pages for hosts.allow, exports, and exportfs. Am I missing something? what is the package name for nfs? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup problem
On Tuesday 11 February 2003 07:51 am, Phil Beder wrote: when I check to status of nfs on the server I';m trying to setup, the rpcinfo -p servername command returns rpcinfo: can't contact portmapper RPC: rompte system error - connection refused the lsmod command shows nfsd, nfs and lockd. what did I do wrong? what am I overlooking. Have you checked to see if the network card has been asigned an IP address and that the network is up? Do su and ifconfig to see. I network is up it should show an IP address asigned to your NIC usually eth0. -- Ken Thompson Payette, Idaho Email: [EMAIL PROTECTED] Linux- Coming Soon To A Desktop Near You Registered Linux User #183936 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup problem
Phil Beder wrote: but I don't seem to have man pages for hosts.allow, exports, and exportfs. man 5 hosts_access will give you the manual page describing hosts.allow (and hosts.deny) man exports man exportfs give me the other man pages. They are part of nfs-kernel-server Am I missing something? what is the package name for nfs? nfs-common1.0.2-1 NFS support files common to client and server nfs-kernel-server 1.0.2-1 Kernel NFS server support Good luck Andreas -- Prof. Dr. Andreas J. Guelzow http://www.math.concordia.ab.ca/aguelzow -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
NFS setup problems
Hi Folks, I have been trying to setup NFS on my system and run into problems. I have been reading the security part of NFS-HOWTO. strings /sbin/portmap | grep hosts shows that my /etc/hosts.deny and /etc/hosts.allow are not being read at all. my /etc/hosts.deny has portmap: ALL lockd : ALL mountd : ALL statd : ALL rquotad: ALL /etc/hosts.allow it has the above interies expect that instead of all it has the ip number of the machine that i want to allow access the nfs share. pcinfo -p info is supposed to show nothing at all according to the doc. however there a bunhc of lines of portmap status, nfs nlockmgr mountd hat are listed when rpcinfo -p is run. I have recompiled the kernel for NFSv3 support and I have nfs-common and nfs-kernel-server install. quota is also installed. I don't know where I am going wrong. I have attached the output of strings /sbin/portmap if that could provide a clue. thank you regards Harshu /lib/ld-linux.so.2 __gmon_start__ libwrap.so.0 _DYNAMIC hosts_ctl allow_severity _init _fini deny_severity _GLOBAL_OFFSET_TABLE_ libnsl.so.1 _IO_stdin_used libc.so.6 authunix_create ioctl strerror errno memcpy svc_sendreply perror dup2 malloc svctcp_create xdr_bytes svcerr_decode socket xdr_void xdr_pmap svcerr_noproc abort fprintf svc_register bind __deregister_frame_info chdir wait3 signal openlog clntudp_create xdr_u_long xdr_opaque sys_nerr fork getopt inet_ntoa getrpcbynumber syslog xdr_pmaplist sys_errlist sprintf stderr exit get_myaddress daemon __libc_start_main open setsid svcudp_create svc_run setuid __register_frame_info close xdr_int free _errno _edata __bss_start _end GLIBC_2.1 GLIBC_2.0 PTRh ,WVS ,WVS WVS usage: %s [-dv] -d: debugging mode -v: verbose logging portmap: fork: %s portmap cannot create udp socket: %m cannot bind udp: %m couldn't do udp_create cannot create tcp socket: %m couldn't do tcp_create run_svc returned unexpectedly %s: %m server: about do a switch svc_sendreply CALLIT (prog %lu): fork: %m setuid(1) failed: %m portmap : request from unauthorized host : request from unprivileged port : request from non-local host : request not forwarded unset null getport dump callit connect from %s to %s(%s)%s portmap: out of memory socket SIOCGIFCONF SIOCGIFFLAGS SIOCGIFADDR cannot find any active local network interfaces /dev/null 0123456789 @(#) Copyright (c) 1990 The Regents of the University of California. All rights reserved. @(#) portmap.c 1.6 96/07/06 23:06:23 @(#) pmap_check.c 1.8 96/07/07 10:49:10 @(#) from_local.c 1.3 96/05/31 15:52:57 Unknown error: -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: NFS setup problems
Hi Folks, I have been trying to setup NFS on my system and run into problems. I have been reading the security part of NFS-HOWTO. however there a bunhc of lines of portmap status, nfs nlockmgr mountd hat are listed when rpcinfo -p is run. I have recompiled the kernel for NFSv3 support and I have nfs-common and nfs-kernel-server install. quota is also installed. I don't understand. Your having problems, and yet rpcinfo -p shows the correct info? If you have full NFS services running, and your system is allowed access, running rpcinfo -p SHOULD show a bunch of stuff. What mount command line are you using, and what is the message that you see(if any) when you try to mount, and what does the log say on the server? I really do not trust the tcp_wrappers(hosts.allow/deny), I don't know why, just paranoia maybe. I much prefer to firewall the ports entirely. I run 2.2.19 and this is what I use: PORTS=`rpcinfo -p | awk '{print $4}' | grep '[0-9]'` for rpcport in $PORTS do /sbin/ipchains -A input -s 0/0 -d 0/0 $rpcport -j REJECT -p tcp -i eth0 /sbin/ipchains -A input -s 0/0 -d 0/0 $rpcport -j REJECT -p udp -i eth0 done then I have the ports blocked again on my exterior firewall. but it sounds like rpcinfo is able to detect the services, so whats the problem? Unless your running rpcinfo from a host which is not listed in hosts.deny, in which case the format of your hosts.deny may be off (service name misspelled or something). I think a firewall would be more effective. nate -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
nfs setup
Hi, This is a problem I tried to solve for about 3 weeks without much progress. My intended system setup has a dependency graph as follows... Sys-A -+--: Sys-B ---: Sys-C | | +-+ Where both Sys-C and Sys-B depends on Sys-A for an administrative directory containing the motd and hosts.* and Sys-B contains the home directory for both Sys-B and Sys-C. and Sys-C is a debian mirror. The reason for such an unusual setup involving Sys-A is that the uptime for Sys-B and Sys-C is dependent on the availability of the systems in my lab but Sys-A is consistently up. My problem is in the setting up of the nfs. Sys-B and Sys-C refuse to startup rpc.nfsd. they return the error: nfsd: could not make a UDP socket What is puzzling is that Sys-A's nfsd is up and running without any problem. I have tried copying the kernel to Sys-B and Sys-C to no avail. The intrigueing thing was that Sys-C actually worked twice.. both when I changed the kernel (when I copied Sys-A's kernel over and when I upgraded Sys-C to Debian 1.2). Unfortunately, it only lasted one reboot... afterwhich nothing else I tried worked. rpcinfo on Sys-(B|C) reports: No remote programs registered. Which is funny because `ps -xaef | grep rpc` returns: 712 ? S 0:00 /usr/sbin/rpc.nfsd INIT_VERSION=sysvinit-2.69 previous=N T 715 ? S 0:00 /usr/sbin/rpc.mountd INIT_VERSION=sysvinit-2.69 previous=N 662 ? S 0:00 /usr/sbin/rpc.portmap INIT_VERSION=sysvinit-2.69 previous= But if rpcinfo is ran locally, they return: program vers proto port 102 tcp111 portmapper 102 udp111 portmapper 132 udp 2049 nfs 132 tcp 2049 nfs 151 udp823 mountd 151 tcp825 mountd Sys-A and Sys-B are now on Debian 1.1, Sys-C is running Debian 1.2 I will upgrade Sys-B to Debian 1.2 when day breaks but am really reluctant to change anything on Sys-A, (unless Sys-B is willing to work). Things I have tried: 1) (Recompiling|borrowing|stealing) kernels from 2.0.(0|6|27) 2) Toyed with different setting of /etc/exports and /etc/fstab 3) Different settings of rpc.nfsd and rpc.mountd and rpc.portmap 4) Reinstalling Debian 1.1 and Debian 1.2 Can anyone please give any suggestion? Thanks Just me, Wire ... -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]