Works with unmanaged 1Gb switch, but throws 'Failed to mount NFS Volume' with managed switch
-
This shows that STP is enabled by default. It’s surprising that you got to the “[B]Fatal Error: Failed to mount NFS Volume”[/B]
Was this during a upload/download or does this occur before the FOG Menu appears? -
From your reply I gather the STP should be turned off. However, in my case the STP is turned off. I also turned off everything else I could find, including all the smart security DOS prevention etc to try and get it as close to an unmanaged switch I can. Most functions are disabled after a factory reset in anyway. Once it is working I could work from there adjusting the switch to for our environment while checking that FOG still works.
But, for the moment FOG is still saying ‘Failed to mount NFS Volume’.
I’m scheduling an upload before PXE booting the machine. It then gets an IP via DHCP server and boots the PXE images etc. There is no menu in this case, but I did boot it to the menu and did a full registration. This worked with no error messages.
Any ideas? (I’m hoping it is just something silly).
Thanks in advance.
-
My guess,
Does the /images/dev/.mntcheck and /images/.mntcheck files exist?
-
/images/dev/.mntcheck - does exist
/images/.mntcheck - does not exist (should I create one?)
FOG Ver: 1.2.0
OS Ver : Debian GNU/Linux 7.7 (wheezy) -
Yes you should create it with:
[code]sudo touch /images/.mntcheck
sudo chmod -R 777 /images[/code] -
OK did this:
touch /images/.mntcheck;
chmod -R 777 /imagesBut it is unchanged. I’ve attached a pic of the screen for you.
[ATTACH=full]1576[/ATTACH][url=“/_imported_xf_attachments/1/1576_2014-12-23 20.23.34 (1).jpg?:”]2014-12-23 20.23.34 (1).jpg[/url]
-
So it never get’s beyond this point and has stopped throwing the “Failed to mount” message?
-
After a while it times out and then throws the ‘Fatal Error: Failed to mount NFS Volume’ and then says it is going to reboot the PC within 1 minute.
FOG works perfectly behind a dumb switch, but this is the behaviour as soon as I put it behind the managed switch. I kept the device to be imaged the same to be sure that it isn’t the device to be cloned that is at fault. I also tried other machines to be imaged but same result. Definitely something to do with the switch it seems.
Creating the /images/.mntcheck didn’t seem to change anything.
One thing though: While setting up FOG I went through a number of steps found on the wiki etc. One of them said to change the FOG Storage Nodes username and password to the standard: fog & password. I kept the fog username and password like this. The username was ‘fogstorage’ and I never knew what the password was. Could I have broken it this way?
Sorry first time installing FOG. I’m quite familiar with Clonezilla and the bouquet of software it comprises. Many of which FOG seems to use too. But, this is the first time I’m trying FOG. I searched your forum for my problem but didn’t find anything similar - so I’m hoping I’m not a noob wasting everyone’s time.
-
[quote=“MrsPotter, post: 40335, member: 27740”]
One thing though: While setting up FOG I went through a number of steps found on the wiki etc. One of them said to change the FOG Storage Nodes username and password to the standard: fog & password. I kept the fog username and password like this. The username was ‘fogstorage’ and I never knew what the password was. Could I have broken it this way?[/quote]I don’t think so, though changing the Storage node username and password may break other functions of fog, it shouldn’t impact imaging. The FOG Username and Password correlate to the FTP user and password (also the local *nix user and password) on the system. Without this File sizes and image uploads may be broken, but from a download standpoint, all should be fine.
Can you ensure that the nfs server is running and that firewall is disabled on the system hosting fog? While I realize it works behind a “dumb” switch but seemingly fails behind a “managed” switch, It still files like a firewall issue more than a switch.
Ubuntu firewall disabling would be:
[code]sudo ufw disable
sudo service nfs-kernel-server restart[/code]Redhat firewall disabled would be:
[code]chkconfig iptables off
service iptables stop
service nfsd restart[/code]Redhat also natively enables SELinux so you’d have to disable it and restart the system if this is the case.
As things seem to work behind a dumb switch, can you just remove the managed switch and verify that this is still the case?
[quote=“MrsPotter, post: 40335, member: 27740”]I searched your forum for my problem but didn’t find anything similar - so I’m hoping I’m not a noob wasting everyone’s time.[/quote]
Regardless, we’re here to attempt to help. If it were something fairly common I might tell you to search and provide a link, but I’m not recalling anything of this type of problem (where it works on dumb, but not on managed).
It’s never a waste of time to ask a question.
-
Hi Tom,
Your help is much appreciated. I haven’t lost interest - it is just the client has closed offices for Xmas & New Years so I currently don’t have access to the hardware for further testing. I’m currently testing FOG as an alternative to Clonezilla to do regular disk clones for backup and archival purposes. Clonezilla worked great up until now, but FOG seemed like an slightly more polished alternative providing a little more automated feedback etc.
So I’m actually mostly interested in the upload (to server) part and in case of emergency the download (to client) bit. I’m having trouble with the upload at the moment. Didn’t get to the download part yet.
I’ll post once I had a chance to fiddle with it - I’m sure I disabled the firewall straight off the bat, but I might have forgotten about SELinux.
In meanwhile - Happy New Year!
MrsPotter.
-
I finally had a chance to look at this project again - I wiped the disk, re-installed a ‘blank’ ubuntu and re-installed FOG. Nothing special, and now it seems to work. Via dumb switch, and via smart switch.
But, thanks for all the help thus far!
