Networking :: Mounting Unique Filesystem Or Directory By Network Booted Host?
Apr 6, 2010
I have a number of diskless hosts which are to bootth PXE and mount their root filesystem on an NFS share. Most of the NFS share will be common to to all of the diskless hosts. No problem with that part.I also need part of the NFS shared filesystem to be unique to each individual host. The only difference between the hosts is the ethernet MAC. I use it in the DHCP server to key a specific unique IP to each host. I would like to be able to use either the MAC, or the IP, or any other unique identifier that can be derived very early in the boot process to access a filesystem or directory that is unique to each diskless host.
I need/want to do this because I need/want to isolate all of the system maintenance to one small part of the boot host. Modifying the hardware (these are VME CPUs) in any way is not an option; they must be field swappable, with all maintenance isolated to the boot host(s).I know I could run ifconfig and parse its output to derive an IP &/or MAC, but that seems a bit kludgey and fragile. The DHCP server is presently assigning IPs. I know I can arrange to pass a unique option value via the DHCP server, but don't know if or how this can be retrieved on the diskless nodes. I am presently passing the 'root=' kernel argument from the bootloader, but I don't think there is any way to pass other filesystem information using that mechanism.
I know I can pass some kernel arguments from the bootloader, but there doesn't seem to be any generic message that can be retrieved in userspace as part of the boot process. I've scanned around in the /proc filesystem, looking for the IP or anything else that seems to uniquely identify the host, but don't see anything there, either.The init process has not yet been cast in stone, but is very likely to be a BusyBox built-in
I am having permissions errors every time I try to mount a windows host. I have a linux server and all the windows computers can see that computer and its files, but we wanted to start backing up the linux machine to one of our other computers. so I tried to mount one of the computers. here is the sequence of events:
Code: $mount -t cifs //192.168.1.194/Admin$ /mnt/Anita-comp password: (I have no password so I left it blank) Mount error (13): Permission Denied I tried all sorts of passwords we use around the office and none of them worked.
I then decided to try mounting one of our other computers. this one looked like it worked fine. no error messages at all. (I left password blank) so I look in my filesystem and the mounted drive is not in the /mnt/Anita-comp file. What gives?
I have two Linux hosts (both running Ubuntu 10.04) and I want one of them to have access to part of the filesystem of the other. I.e. I want remote1 to be able to access the files of remote2:/home/username
This is simply to allow me to access my home directory on the other machine, read-write, and there's no need for other users to be able to access this too. The used IDs for 'username' are different on each host, for reasons I won't go into (Likewise, ADS, etc).
I tried Samba but I found it unreliable. For example, I have a program that monitors a directory and updates a list of around 12,000 files. I found that this list was constantly changing with files dropping out and some files being randomly unreadable. It was too unreliable.
I had something working quite well on remote1 with:
This works perfectly for a while. Unfortunately remote1 is a laptop and is regularly disconnected from the LAN. This seems to completely break things when this sshfs filesystem is mounted, and typically requires a reboot to recover. I started to look into NFS but I ran into problems with this a few years ago with regards to user id mappings. At the time, I found there was a user-space NFS server that dealt with this but it had a bunch of its own problems. But that was NFSv3. Can NFSv4 'kernel server' handle different user ids and map between them properly? I really don't want to use NIS, and I can't actually change the user IDs anyway - they are set by external means.
Also, how does NFS handle an 'unreliable' network connection? Will it automatically reconnect or will the share be broken until it is remounted? What about something like CIFS? Can this cope with the network being regularly and unpredictably broken? This might have the same problems that Samba did, but it's worth a try if it will be suitable.
I am looking for c/c++ code (no third party lib dependency) to get the physical address of ethernet card on mother board.Most of them suggests to use the linux commands and process the output. For example "/sbin/ifconfig eth0".The problem with above command is that when you run this on a linux variant which is running using VirtualBox on windows host, it's not returning the correct mac address. In this case the correct mac address is the one set for individual virtual machine."/sbin/ifconfig eth0" works fine when linux variant is installed normally.There is another case when linux variant is installed normally and if you install virtualization apps such as VirtualBox or VmPlayer, they create a virtual adepter for their own use. In that case running "/sbin/ifconfig" returns all the adepters. (physical and virtual)
I am currently running Jaunty 9.04 and find the xorg.conf in /etc/X11. I just booted Lucid 10.04 from the Live CD just to take a look at things. I could no longer find xorg.conf any place in the booted file system.
I am running VMware workstation on a Linux host and WinXP guest using NAT. I have shared a folder on WinXP, and I am able to mount it on the linux host with: mount -t smbfs -o username=Administrator,password=mypasswd //192.168.100.129/Shared /mnt/tmp This works fine. However, I want to mount the same folder from a different linux machine (not the host).
I added port forwarding in /etc/vmware/vmnet8/nat/nat.conf: [incomingtcp] #445 = 192.168.100.129:445
If I run: mount -t smbfs -o username=Administrator,password=mypasswd //external_ip_number/Shared /mnt/tmp
I am getting: mount error 112 = Host is down Refer to the mount.cifs(8) manual page (e.g.man mount.cifs)
It doesn't work even if I try the above command from the linux host itself. It only works if I use the internal IP number. I suspect vmware is not forwarding port 445 but I am not sure how to verify it. I am not running smbd on the linux host, in case that make any difference. I didn't think that was required because the server is really on the WinXP and the linux is just a client.
The internal network is behind nat done by the PC Router.The TP Link is recieving wireless signal from outdoors and it has switching and basic routing capabilities. I'm using the PC router for better routing options.PC Router (or R for short) is a triple-booting machine - Linux, FreeBSD and Windows. It has two lan cards - external (ext_if) - 100Mbps Realtek 8139 and internal (int_if) - 1Gbps integrated Realtek 8169.The problem is that all traffic from R to the network is slow - about 5-20K, while the traffic in the oppoiste direction is all right - about 10MB that is fine for 100Mbps cables, NICs and switches. The problem persist no matter the OS the pc R is running.I've tried some debugging on the situation as follows:
- put another PC at the place of R - everything is fine. That exclude the possibility of damaged cables, RJ-45s, switches and etc. - connected both of the NICs to the Internet while the internal network is being disconnected and they both work fine (no delays) - traffic shaping is not running - there is nothing in firewalls except NATing the internal network (and it is working fine). Actually these firewall rules have been operational for more than months and everything was fine untill a week or two ago. - changed the internal NIC with another - connected the internal network directly to the TP and all of the PCs are getting good network performance. Then connected the R machine to the TP as well and there was good performance between the internal network PCs and R. - R has good performance to the TP. In fact everything has good performance directly to the TP (when not connecting trough R). - the problem persist only between R and machines from the internal network.
I have an interesting problem I have had some troubleinding answers to with researching. The answer is probably so obvious that I should be able to see it. Here is my prob.I am trying to mount a root directory onto the same system (just for testing purposes)I am using nfs and have nfs-kernal-server nfs-common and portmap installed.if I have this in my /etc/exports
Code: / 192.168.1.0/24(rw,no_root_squash,async) then do
I am trying to get a ubuntu machine (client) onto a windows active directory (domain) this i have done and you can login using winbind to the client desktop no problems however i dont want the domain users' home directories on the client machine so i have set up a ubuntu server (samba) to hold the home directories now so far i have been able to set up a share which both windows and linux can read and write to with no passwords needed, and if i have modified the /etc/fstab file on the client to mount the samba share on startup however if i login as a domain user it fails to create a home directory on the share with the following error:
"/mnt/home/admin2" does not exist
now the main question is this in fstab i have used a cred file stored in /usr/share/.smbcred which should be accessible by all users right? i know i can put the creds into the fstab file but i can't find how and it is just failing to mount when i try so if you know how i will try that, also does fstab run before or after the home directory is created as if it runs after then the cred file is working but the home directory is looking at a location which hasnt been mounted yet or if fstab runs before then the location is not mounting right (hoping for second one )
I have a computer (C1) to which I can connect through the Internet (ssh, for instance) (it has a static ip and though it is sitting behind a router, the appropriate ports are all forwarded). I have another computer (C2) that doesn't have a fixed ip address and sits behind a router that I cannot fiddle with (so no port forwarding here). I would like to know if there is any way to connect from C2 to C1 such that a directory on C2 would be mounted on C1.
I have attached a 2 TB LUN to the HP Blade Server running CentOS 5.3 via a Qlogic HBA. To provide the cluster users with sufficient storage areaa, I want to move only the home directory from the default partitioning schema to the attached storage and leave the remaining partitions on the main harddrive of the server. So, having copied user directories & files to the new location, i.e., /cluster/home2 on the new storage partition, what modification(s) should I make on the server? /export/home is the default location for the users. /cluster/home2 is NFS shared directory serving as the new home location for the users.
today I tried to configure a network route to a host for testing my network interface. Code: route add 192.168.1.15 eth0 As I have to eth interfaces and both interface got their IP from DHCP (192.168.1.11 and 192.168.1.12) and are in the same subnet, I shut the other interface down:
Code: ifconfig eth1 down Then I tried to test the interface by doing a ping to 192.168.1.15. Problem: When I unplug the cable from eth0 (and eth1 is still plugged) the ping still works. Somehow my linux (it's debian) powers up again eth0 and pings over this port.
How can I stop my linux doing this. I just want to have the route added only on the one interface - not the other. Is it maybe some case of a default-gateway?
Here is an overview of my system, consisting of two computers, a D-Link router and a ZyTel 660 DSL modem/router. Computer One (Comp1) is dual booting XP and Ubuntu 8.04. Computer Two (Comp2) is booting only Ubuntu 8.04. For years, I had been running and networking both of these computers with XP. Both have been using static IPs. In order to have Static IPs and forward certain ports, I had to configure the D-Link router as an Access Point and use the ZyTel as the defacto router for the network.
The Static IP info for the two computers is: Comp1 IP ADR 192.168.2.20
i am getting weird results after mounting the /home directory from the server. seems like the client sees the ldapuser and it's directory from the server side.. /home/ldapuser. but it's pulling the "blank" user as it's permission. blank is the actual user on the client side. any ideas why?
server side: [root@localhost ~]# showmount -e Export list for localhost.localdomain:
I've mucked through and figured out how to mount a windows share. I can access the folders I was looking for, but the windows share was not what I thought it would be. I was looking for the specific shared folder. Instead I got a root level parent directory that included the folder I wanted, and a couple others.
smbclient -L <ipaddress> gives me a parent directory on the root
First question: Can I mount a specific folder within a share? Second question: Could somebody define share? I thought it was the specific shared folder, but that doesn't seem to be the case.
I have Centos installed and runing Vmware server inside, inside the Vmware server i have two V MAchines runing, how can i let those machines run outside the host ( mean listen to other computers in the outside network, i am after proxy server.
Following a slight dose of vandalism, I have invested in a security camera system.It is basically a hard drive in a box which connects to a couple of cameras.There is also a USB lead to connect it to a PC, and some software that enables Windows to look at the recordings (I can't get it to work on WINE).I was hoping to simply bypass that, and just look at the drive directly, but it seems from here that the disk "doesn't contain a valid partition table".Obviously, the box thinks it has a valid partition table, else it would not work.I was quite surprised: I had assumed the box would have some kind of cut down Linux based system, but if that were the case, you would expect it to use a recognisable filesystem.
My Windows guests see my VMware Host Only folders as \.hostShared Folders But when I try to see them from a Linux guest, all I get is frustrated. No permutation of that seems to work. Because the Windows guests see the shared folders I know the host is doing its part in providing them. After two days of not finding an answer on the VMware Workstation forum I finally figured out that the problem really is how to look for them with Linux. With Linux I can turn off all NICs accept the Host Only NIC, put smb://172.16.138.1 into the address bar and see all the shares on the host, but not the folder designated as the host only folder.
I have a somewhat complicated network setup that I am testing on an internal network. I have the following route setup:
PC Client (192.168.2.100) --> Router (192.168.2.1) --> DSLAM (on our internal network) --> PPPOE Server (192.168.9.1) on Linux Ubuntu 8.04 on interface card eth2.
On the same Linux Ubuntu Machine on interface card eth1 (static IP 192.168.5.100), I have an Asterisk SIP server plugged into it. SIP Server = 192.168.5.101 (static) I need the PC client (192.168.2.100) to register via SIP soft phone to register on the SIP server (192.168.5.101). From the PC client I can ping as far as the eth1 interface (192.168.5.100)...but cannot ping the SIP Server (.101).
From the SIP server (18.104.22.168)...I can ping 192.168.5.100, I can ping 169.254.5.228 (Eth2:avahi), but not sure what that is. I cannot ping the pppoe default gateway (192.168.9.1), which I think you cannot anyway. No firewalls are running. My IPtables I cleaned out totally. I think it might be as simple as a route add, but I really have no clue. Tried building a virtual bridge using brctl LINUX betweeh eth1 and eth2, but that made things worse (could not ping anything after that)
I am trying to configure MPD (music player daemon) to work on my headless ubuntu server. Everything works well, but MPD cannot see the music files stored on my windows XP main computer. These files are stored at smb://LASTNAME/share/music Is there a way to directly mount this drive, so it is accessible at something like /mnt/music, in order that they work with MPD?
I have 2 Ubuntu machines: a desktop in my bedroom, and a laptop.
I have my music shared on my desktop machine, and can access it through the network menu item in the nautilus manager, but I want the files from the share to be mounted on the disk so I can access it through the commandline.
If I right-click the shared folder in Nautilus, it says its location is smb://rob/music
If I do:
mount -t smbfs //rob/music /mnt/music, it tells me that it cant locate rob.
So I try "ping rob" and that doesn't work.
I can't make a hosts entry for rob which happens at this moment to be 192.168.0.8, because my router assigns different IP addresses to various machines at different times, and I cant seem to find a way to make static maps from MAC address to ip address.
So, how come nautilus can see my samba share on the machine "rob", but the mount commands cannot?
I have a network PC running Win7 that u use for storage of all my media; movies, music and pictures.I can connect and use the share just fine using the "connect to server" option under places menu.I think i need to modify the fstab file but I am not familiar enough with it to do this.Have searched other threads for help but I am doing something wrong.HP Laptop running Ubuntu 10.10 connecting to a win7 share through a router.
I am using Ubuntu 9.10 (along with Win xp). I have to authenticate everytime I mount filesystem (My d: , e: drives in windows). Or to connect to the internet (I use mobile broadband) I have to authenticate, also if I have to install something from synaptics I have to authenticate. I know this is good for security but I am the only person using my computer , so is there any way out of this authentication business.
I am new to Linux platform and have installed RHEL 6 on my computer. I have gone through different posts on websites regarding how to mount windows drive on to the Linux file system.Here are few results of my trial where results are slightly overwhelming for a guy like me:
1. Mounting NTFS file system using mount commandAs a su-) " $mount -t ntfs /dev/sda1 /media/MyFilesystem"
Output: error: unrecognised file system ntfs
2. I followed step by step instructions of installing Fuse and ntfs-3g_ntfsprogs but had no luck:
while using this command for the installation of for FUSE-0.9(which I guess prepares files for installation) I am getting this error : " Command use: ./configure --exec-prefix=/; make;make install"
Error: "Checking Kernel source directory.../usr.src/kernels/2.6.32-71.e16.x86_64...Checking Kernel source version.......Not found " Configure error: cannot determine the version of the Linux kernel source.Please configure the kernel before running this script.
Woooh , as u can see my Linux version is mentioned in that message so can anyone lemme know whether this issue is because of the incompatibility of Fuse with my version of Linux or am I suppose to make some changes in the command itself.
20 packets transmitted, 0 received, 100% packet loss, time 19152ms However, ping works fine if I use IP addresses instead. I would put the address for "file-server" in my /etc/hosts file, except for the fact that DHCP changes it's address occasionally.I attempted changing the line
send host-name "<host-name>"; in /etc/dhcp3/dhclient.conf to send host-name "file-server";