mounting /images failed: Connection timed out

[george1421]

@Tom-Elliott for reference this is what my fog server has (note the fsid is different for each export [share]):

pi@pi01:~ $ cat /etc/exports 
/images *(ro,sync,no_wdelay,no_subtree_check,insecure_locks,no_root_squash,insecure,fsid=0)
/images/dev *(rw,async,no_wdelay,no_subtree_check,no_root_squash,insecure,fsid=1)

[sjensen]

@george1421 What does that affect? Should I change it?

[Tom Elliott]

@sjensen I need to see it.

/images should have FSID of 0
/images/dev should have FSID of 1.

If you must change these values for other reasons, they cannot be the same, and they should be “iterative”

[george1421]

@sjensen said in mounting /images failed: Connection timed out:

@george1421 What does that affect? Should I change it?

In a word, please show us your cards first. Then we will tell you if you won or not.

[sjensen]

@george1421

[Tom Elliott]

Please give a shot at changing the fsid’s so they are in order:
/images with fsid 0
/images/dev with fsid 1.

Also, please give a try at:

sudo touch /images/.mntcheck
sudo touch /images/dev/.mntcheck
sudo chown -R fog:root /images
sudo chmod -R 777 /images
sudo exportfs -a
sudo service nfs-kernel-server restart
sudo service ufw disable

[sjensen]

@Tom-Elliott Ok I did what you asked. Should I try to upload an image again?

[Tom Elliott]

Yes please.

[sjensen]

@Tom-Elliott same result…sorry

[sjensen]

@Tom-Elliott yesterday I did a kernel update would that have broken anything? I did it via terminal.

[Tom Elliott]

@sjensen The fact that it’s saying timeout leads me to think the the connection is unable to reach the server.

[george1421]

@Tom-Elliott Possibly schedule a debug deploy and then check what FOS is seeing?

[Tom Elliott]

@george1421 Sure? I don’t know. This seems awefully fishy and we aren’t getting anywhere.

The stuff all points that something else is blocking the connection. Maybe the subnet the systems are booting from (and being issued IP addresses) is not allowed to request data?

[george1421]

@sjensen said in mounting /images failed: Connection timed out:

@Tom-Elliott yesterday I did a kernel update would that have broken anything? I did it via terminal.

I guess we need to collect a little more information here then.

What version of FOG are you using?
What version of kernel did you just update to?
Why did you update it?
What hardware are you trying to capture? (manufacturer and model)
Is the FOG server and target computer on the same subnet? (I think you already answered this)

What you are experiencing is not typical with FOG. We need to identify where the gap is. It appears the target computer has network access, but it doesn’t have nfs access to the fog server(??). You can mount the nfs shares locally on the fog server, so the fos engine should be able to do the same. You have confirmed that the IP address of the FOG server matches the IP address from the error page. There is something going on here that is for sure.

[sjensen]

@Tom-Elliott I dont think this makes a difference, but my fog server is a virtual it is on one vlan and my host is on another vlan. I can connect to the web interface for Fog with no issues from my computer which is on another vlan.

[sjensen]

@george1421 Fog version 1.2.0
the latest kernel (i saved the old kernels)
I updated because i was trying to download an image to a computer (HP prodesk 600 g2) and was getting an error “tsc fast tsc calibration failed”
all posts I read said to update the kernel.
I am trying to capture a HP 6005 usff. I have made a test image of this pc prior.
Yes both Fog server and client are on the same subnet.

[Tom Elliott]

@sjensen Your information is conflicting.

You state the FOG Server and the client are on the same subnet, but the post before that you state they’re on different VLAN’s.

Now, being issued different IP’s is fine as your DHCP controller can handle passing the boot files and with the use of IP Helpers all will seem to work fine.

The difference, however, is something is obviously not right and we now have conflicting information making it that much harder to figure out.

If they are on separate vlan’s I’m assuming there’s a switch handing out information and there’s probably a central point where all things reach out through. After that, there’s probably a firewall and traffic is all routed to that point?

I’m just trying to understand the layout. Two clients on the same subnet (Server and host) should have no problems mounting an NFS share. But if you throw a firewall into the mix and separate vlan’s, I can only guess the network has no idea where to go, or it’s entirely blocked from that particular connection.

[george1421]

@sjensen just for clarity (since you are using FOG 1.2.0) what is “the latest kernel, specifically” The latest kernel for FOG 1.2.0 is not he latest kernel for FOG 1.3.0. If you revert to the previous kernel and do a compatibility test from the fog menu, do both networking and hard disk pass?

Also for clarity has this setup ever worked? I know you said that it worked the other day, but how much history do you have with this FOG server?

As Tom noted, and I also picked up on. You talk about vlans and then mentioned that both the fog server and the target computer is on the same subnet. I see a discrepancy here. I specifically asked if they where on the same subnet to ensure there was no kind of screening / firewall router between the target computer and the FOG server that may be blocking NFS traffic, but passing http traffic.

You mentioned that FOG worked before (for some system), does it sill work if you tested with that model?

As noted before this system is acting inconsistent with what we would expect from FOG. There has to be something that changed beyond the kernel.

We may need you to do a debug capture for this computer. (the debug capture for 1.2.0 is on the advanced menu in the host settings page). By doing a debug capture and then pxe booting the target computer FOG will drop you to a command prompt on the target computer so we can do some additional debugging steps.

The first thing I would try is for the target computer (running in debug capture mode) to see if it can ping the fog server.

[sjensen]

@george1421Please understand I am very new to Linux, I appreciate all the commands you have sent. I downloaded the latest kernel for 1.2.0, however i do not remember the version. I can try to revert to old kernel.

To clarify, yes the setup worked. I was able to upload and download an image. Host had an ip address ending in 6.62, Fog server Vm ip address ending in 28.10, client got its ip from the dhcp server in the subnet as fog server 28.xx.

No the original system does not work any longer.

TO my knowledge I have not changed anything else except the kernel.

As a test this morning I put host, server, and client on the same subnet. I still get the same result. I am thinking its got to be the kernel. Thoughts?

[Tom Elliott]

@sjensen The kernel has nothing to do with connecting to the NFS Mount point. Support for NFS is embedded in every kernel in the same way (so if one kernel works all the rest will as well).

You can try reverting to the original kernel but I don’t think that will fix the issue either. From a client within a Debug mode: (From FOG GUI goto Host->Basic Tasks->Advanced->Download - Debug)

Let the system boot and press enter if/when prompted to. It will drop the client machine booting into a console shell.

Let me know when it’s there.