Network Connections to random IPs fails?

Status
Not open for further replies.

I-Tech

Dabbler
Joined
Aug 14, 2015
Messages
36
Hi..
Have been running v9.3.1 for a while now (latest updates last month or so), and have just had an issue with network connectivity.

System was running for weeks, no issues, and then one of the backups to the NAS fails (the FreeNAS box is a backup target) The system has been working very stable with good throughput from backup sources.

From the server backup error log, the backup target is unreachable.

From the FreeNAS system, I log into ssh, cannot ping gateway at 10.0.1.1, can ping server at .2, can ping other systems at .7 .19 .22, but cannot ping .20 or .21.. seemingly random systems.

As the system was idle at the time I was checking, did a quick disconnect /reconnect of network cables (set up in lagg config), same results.
did a quick reboot.. same results.

some vague memory of my testing 9.2 system earlier last year brought IPv6 to mind, so I disabled that.. and pings worked again. All systems were now reachable.

As I recall having done this before, I also recall switching IPv6 back on for some other reason, which eludes me at present (will skim through my notes and see if I can dig up more details) and I also recall disabling lagg after disconnecting one of the network cables (bce0, bce1) and having pings work again.

Has anyone else had any similar issues?

There have been no changes immediately preceding the failure that I can see having any effect like this.

All Servers and gateway are connected to the same switch.

thanks in advance
 

SweetAndLow

Sweet'NASty
Joined
Nov 6, 2013
Messages
6,421
Don't set up lagg unless you know what you are doing and have a switch that supports it. Ipv6 should work just fine if your router supports it. Ping also doesn't care about ipv6 because it only does ipv4 address. Ping6 is for ipv6 addresses. If you do a trace route to the address that don't respond to ping what do you get?
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
To be fair, sweetandlow, "lagg" doesn't necessary imply LACP. It's LACP that requires the special gear, in most cases. I have lagg configured right now for simple failover with shit equipment and it's fine.
 

I-Tech

Dabbler
Joined
Aug 14, 2015
Messages
36
Lagg is config'd with LACP and as I said .. appears to (have) work(ed) as expected (for several weeks) .. have tested multiple simultaneous transfers showing gigabit traffic over both network interfaces. if it was going to 'not work' as advertised, I would expect it to 'not work' immediately or shortly after plugging in. have read lots of posts here about this before setting things up.

The switch (TRENDnet TEG-160WS) supports (spec sheet says: Static Link Aggregation / 802.3ad Dynamic LACP) and is set up with this and trunks to another switch (TRENDnet TEG-448WS) and works as expected.

Perhaps toggling the 'Auto configure IPv6' box off resets something on the interface so that things start working again.. as I can switch it back on and the network access continues to operate.

I didn't do a traceroute as the other destinations are on the same switch, well.. if I was to try 8.8.8.8 it wouldn't get past the first hop, as the gateway itself was unreachable at the time.
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630

I-Tech

Dabbler
Joined
Aug 14, 2015
Messages
36
Apologies for the misunderstanding..

the intended meaning was as follows:

"Lagg is config'd with LACP." (in response to your comments)

next context:

"and as I said .. appears to (have) work(ed) as expected (for several weeks)" (meaning as I said earlier, that it appeared to have been working)
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
Do your problems go away if you turn off the LACP, and go to a simple, single, port? You should at least test that hypothesis before we move forward with more troubleshooting.
 

I-Tech

Dabbler
Joined
Aug 14, 2015
Messages
36
I ran the system on a single port for a few weeks during setup and evaluation, ran several data transfer jobs, read up on the LACP Lagg in the forums, configured the system with both ports, ran more tests over a few weeks, everything appeared to be working fine.

I can disable lagg and use a single port again, and I can schedule backups to run at staggered times, so won't have any performance issues.
 

DrKK

FreeNAS Generalissimo
Joined
Oct 15, 2013
Messages
3,630
Sir.

I am merely asking if, as a troubleshooting step, you would please check to see if your problems go away when you revert to single NIC usage with no LACP.

The answer will either be "yes" or "no", and then we can take it from there.
 

I-Tech

Dabbler
Joined
Aug 14, 2015
Messages
36
Ok. will disable lagg / lacp and run on a single interface for a while to see what happens
 
Status
Not open for further replies.
Top