Mounting file system: Failed. Uploading my first image

[Miguel Palacios]

@george1421 here it’s

[Miguel Palacios]

@Quazz the result for rpcinfo -p

[george1421]

@Miguel-Palacios said in Mounting file system: Failed. Uploading my first image:

@george1421 here it’s

Sorry for being stubborn here, but what you are seeing is not logical. I don’t doubt it is happening, its just confusing.

[Miguel Palacios]

@george1421 I am in the same situation…

Say you that the FOG server is installed in a virtual machine on VMware ESXi 5.5.0 (3116895)… I don’t know if this detail shed light at problem…

[george1421]

@Miguel-Palacios I had my fog servers on centos 6.5 running on vmware 5.5 for over a year. This Jan I rebuilt the server using centos 7 and the current fog trunk. Both worked without issue. Now that I think about it, I have a dev fog server that is still running centos 6.5/6.7. I may need to power it on so we can compare services or what ever.

[Miguel Palacios]

@Sebastian-Roth I attached the result for our test.

0_1459517616214_nfs.pcap

[Sebastian Roth]

The PCAP file is showing a perfect connect form client to server on port 111 (portmap) to get the NFS port. Then… silence. After some more digging and testing it turns out that an intermediate cisco switch is causing the problem!

[george1421]

@Sebastian-Roth Probable, but unbelievable at the same time unless that cisco switch was doing some type of acl filtering this should not normally happen. This results make the first one for me at least…

Great find because I was lost on this one.

[Wayne Workman]

@Sebastian-Roth said in Mounting file system: Failed. Uploading my first image:

The PCAP file is showing a perfect connect form client to server on port 111 (portmap) to get the NFS port. Then… silence. After some more digging and testing it turns out that an intermediate cisco switch is causing the problem!

Can you give more details on exactly what was wrong with the switch?

[Miguel Palacios]

Hi guys!
Thanks to the great help of @Sebastian-Roth we have managed to figure out the problem. This is a switch CISCO Small Business that had configured between the main switch and the client. I still have outstanding figure out the problem of switch CISCO because it is very rare, there are no complex configurations. If I find something I keep them informed…

Connecting the client directly to the main switch no problem.

Thanks everyone for your help!

See you later!

[Wayne Workman]

@Miguel-Palacios Please give us the model and version of the middle problem switch please?

[Quazz]

@Miguel-Palacios For some people updating the firmware helps, if that’s a possibility at all it might be worth giving it a go.

[Miguel Palacios]

@Wayne-Workman Yes, this is a CISCO Small Business SG 200-08 (8-Port Gigabit Smart Switch). Version 1.0.6.2.

[Wayne Workman]

@Miguel-Palacios said in Mounting file system: Failed. Uploading my first image:

@Wayne-Workman Yes, this is a CISCO Small Business SG 200-08 (8-Port Gigabit Smart Switch). Version 1.0.6.2.

For the record, I have a SG100-D8 in the basement and it works fine with FOG. I also have a SD2005 5 port that also works fine.

I’ve got a newer (metal case) Cisco Small business 8 port gig switch at work that also works, it’s the newer style. Don’t know the exact model because I’m not in today.

[Miguel Palacios]

@Wayne-Workman Yes… I also have other Cisco Small Business (same model) working with FOG but not yet know why some fail sometimes very silly things. It is strange and worth investigating.

[Wayne Workman]

@Miguel-Palacios To be honest, I really doubt it’s the switch… You should look intensely close at the patch cables that were being used as an uplink. Even a slightly loose connection can cause issues, a slightly deformed RJ-45 connector, slightly oxidized connectors, slightly loose connection at the main switch, or just a kink in the cable somewhere… It could even be interference on the patch cable itself, is it running parallel to power cables? That would cause major interference. Keep your patch cables away from power cables!

I really doubt it’s the unmanaged switches fault, and I’m willing to lay 99% odds it’s a cable.

[george1421]

For clarity too. In my deployment area I have a LinksSys/Cisco SLM2008 switch (which I think has become the SG200-08) connected to our 2960 that deploys without issue for FOG.

[Edit] For completeness my SLM2008 is running firmware 2.0.0.10 AND in the FWIW bucket. If I go to CDW and key in SG200-08 the Cisco part number is SLM2008T-NA. Meaning that my suspicion is correct the SLM2008 has been renamed to SG200. [/Edit]

[Sebastian Roth]

@Wayne-Workman said:

I really doubt it’s the unmanaged switches fault, and I’m willing to lay 99% odds it’s a cable.

While I totally agree on what you said and fully trust in your network knowledge (I mean that!) this is not a cable issue from my point of view - not this time at least:
http://www.viktorious.nl/2013/11/05/cisco-sg200-08-nfs/
https://supportforums.cisco.com/discussion/11755791/sg200-08-firmware-issue

Other than that I found this as well: http://serverfault.com/questions/367107/cant-mount-nfs-share-over-tcp

Turns out there was a “security” feature enabled on our PowerConnect switch that took offense to NFS SYN packets with source ports < 1024 (dos-control tcpflag). Suffice it to say, disabling the feature solved the issue.

Reading this again I think this is actually the case here. Why? Because I made Miguel try telnet <fog-server-ip> 56557 (the actual NFS data port) and it worked - as telnet is not using a high source port I suppose.