[lug] NFS on Debian Jessie

[Gary Hodges - NOAA Affiliate]

On 12/07/2016 01:03 PM, Gary Hodges - NOAA Affiliate wrote:
> On 12/07/2016 10:42 AM, Zan Lynx wrote:
>> On 12/07/2016 10:00 AM, Gary Hodges - NOAA Affiliate wrote:
>>
>>>
>>> user at machine:~$ ps -ef |grep statd
>>> statd      401     1  0 09:57 ?        00:00:00 /sbin/rpc.statd
>>> statd      886     1  0 09:57 ?        00:00:00 rpc.statd --no-notify
>>>
>>> So it seems all should be good, but it's not.
>>
>> Test with the firewall on the NFS server completely disabled. Make sure
>> that its iptables rules are all flushed and set to default ALLOW.
>>
>> If that works, it means you'll need to find the documentation for Debian
>> (I found it for Redhat but that won't help you) for how to force the
>> service ports to fixed port numbers and how to set persistent firewall
>> rules to allow those ports.
>>
>> NFS was designed in an age before firewalls and by default its RPC
>> system uses random port numbers which are picked by the port mapper.
>> This was also an age before every service ran on port 80.
>
> IT believes it highly unlikely the issue is on the server end.  They
> have tested three other machines without issue.  That would make sense
> since my desktop machine, also Deb Jessie on the same subnet, worked
> out-of-the-box.

Imagoober and Zan was on to something.  It was a hosts.deny hosts.allow 
issue on my machine.  One must explicitly allow rpcbind in hosts.allow. 
Thanks to all for your patience.  I did pick up some info while going 
toe-to-toe with nfs.  I learned a bit about update-rc.d and was able to 
solve a lingering issue that had been bugging me.

Gary