Just as a side note has the developer quit supporting unRAID? I've had mine (4.4) almost a year there hasn't been an update other then a few betas.

according to the wiki there have been about 20 releases (if you count the beta versions) between 4.4 and 4.5.3

http://lime-technology.com/wiki/index.php?title=Release_Notes

 

I don't think you are looking in the Announcements thread.  A lot has gone on past the "few" betas.  Tom has been focusing more on the hardware for the past few months, but we know as of yesterday he will be putting out a bug-fix release shortly.

 

Tom is currently working on a bug on 4.5.3 that shows all disks as un-formatted when first starting the array.   "Stopping" and "Starting" it gets the disks in sync so they show proplerly.  There is some danger if you press the "Format" button when the display shows the disks as un-formatted, since unRAID will do as you ask (and format your data disks).  

 

As long as you know that the bug exists, and don't press "Format when all our disks are showing as un-formatted," the 4.5.3 release made significant performance improvements to where some people with fast (7200 RPM) disks are reporting 35 to 40MB/s write speeds to the array with no cache disk installed.

 

Joe L.

You'll have to disable the old one in BIOS...

There is nothing that I can find in the BIOS about the Ethernet port. The most obvious place (I believe) would be under Advanced Settings.

 

Under Advanced Settings I have:

 

BIOS

Processor & Clock

Advanced Chipset

IDE Config

PCI/PnP Config (under it it has Local Onboard LAN [disabled])

SuperIO Config

Hardware Health Config

 

So I installed anyway and booted up. It doesn't read from it.

 

 

It is a plain-text file.  Edit it with any text editor, just before you shutdown the server for the upgrade.

Look for this line:  HWADDR=00:12:C4:13:A9:F2

Remove the numbers (whatever they are), and leave it like this:  HWADDR=

 

 

I did this and now after I booted up with the new Ethernet card it didn't work.  I changed the cable back to the on board Ethernet and it working.  However my HWADDR= setting is still blank.

It is a plain-text file.  Edit it with any text editor, just before you shutdown the server for the upgrade.

Look for this line:  HWADDR=00:12:C4:13:A9:F2

Remove the numbers (whatever they are), and leave it like this:  HWADDR=

 

 

I did this and now after I booted up with the new Ethernet card it didn't work. 

I changed the cable back to the on board Ethernet and it working.  However my HWADDR= setting is still blank.

That means that your onboard ethernet adapter is not disabled. Apparently unRAID sees it as working, and is using it.

Put the MAC address of the new card in that "HWADDR=" line, then reboot your server. UnRAID will then look for, and use that new NIC.

 

To see the MAC addresses of all available network adapters, type this command:

ifconfig -a

 

That means that your onboard ethernet adapter is not disabled. Apparently unRAID sees it as working, and is using it.

Put the MAC address of the new card in that "HWADDR=" line, then reboot your server. UnRAID will then look for, and use that new NIC.

 

To see the MAC addresses of all available network adapters, type this command:

ifconfig -a

 

Sorry to report but the HWADDR= setting doesn't make any difference no matter what is there (even blank). It still works off the MB ethernet connection (eth0) and not the card (eth1).

 

Here is my info:

eth0 00:30:48:b0:82:a4

 

eth1 00:1b:21:5d:30:dd

 

No matter which value is in HWADDR= (or no value for that matter), no mater how many times I reboot/reinstall card, or switch the cable the eth0 is always working and the eth1 is never working. There must be another file or system that controls the ethernet port since HWADDR= doesn't.  I'm currently connected to the Tower through eth0 (00:30:48:b0:82:a4) and here is a pic of my network.cfg file to prove it.

 

 

 

EDIT: To make it even more confusing it now has both (eth0 & eth1) listed as the same 00:1b:21:5d:30:dd. So in effect it looks like unRAID changes the hardware to fit the number I put in the HWADDR= field.  Of course only eth0 has an IP address and send/receive info.

That means that your onboard ethernet adapter is not disabled. Apparently unRAID sees it as working, and is using it.

Put the MAC address of the new card in that "HWADDR=" line, then reboot your server. UnRAID will then look for, and use that new NIC.

 

To see the MAC addresses of all available network adapters, type this command:

ifconfig -a

 

Sorry to report but the HWADDR= setting doesn't make any difference no matter what is there (even blank). It still works off the MB ethernet connection (eth0) and not the card (eth1).

 

Here is my info:

eth0 00:30:48:b0:82:a4

 

eth1 00:1b:21:5d:30:dd

 

No matter which value is in HWADDR= (or no value for that matter), no mater how many times I reboot/reinstall card, or switch the cable the eth0 is always working and the eth1 is never working. There must be another file or system that controls the ethernet port since HWADDR= doesn't.  I'm currently connected to the Tower through eth0 (00:30:48:b0:82:a4) and here is a pic of my network.cfg file to prove it.

 

 

 

EDIT: To make it even more confusing it now has both (eth0 & eth1) listed as the same 00:1b:21:5d:30:dd. So in effect it looks like unRAID changes the hardware to fit the number I put in the HWADDR= field.  Of course only eth0 has an IP address and send/receive info.

Yes, it allows you to override the normal MAC address of the hardware.  Not useful for you, and very BAD if you happened to have both LAN connectors attached to your router.  They'd both have the same MAC address and both be trying to answer to the same packets.

 

Joe L. 

Yes, it allows you to override the normal MAC address of the hardware.  Not useful for you, and very BAD if you happened to have both LAN connectors attached to your router.  They'd both have the same MAC address and both be trying to answer to the same packets.

 

Joe L. 

Not sure I understand. What do I need to do to make eth1 my active port and take eth0 offline since purko's method doesn't work?  I only have one Ethernet cable so only one port can be connected at a time.  If I connect to eth1 I get no connection.

EDIT: To make it even more confusing it now has both (eth0 & eth1) listed as the same 00:1b:21:5d:30:dd. So in effect it looks like unRAID changes the hardware to fit the number I put in the HWADDR= field.  Of course only eth0 has an IP address and send/receive info.

OK, I looked a little closer into it, and it turns out that the purpose of HWADDR is not to tell unRAID which NIC to use.

The purpose of HWADDR is to overrule the card's hardware MAC address.  So HWADDR is not what we need.

Just clear the HWADDR line in network.cfg and forget about it.

 

If you have no way to disable the onboard NIC in BIOS, then there's only one thing left you could do:

Assign the onboard NIC to some unused subnet, bring the second NIC up manually (in your 'go' script),

and assign the second NIC a static IP address in the subnet that you are actually using.

 

/boot/config/network.cfg

USE_DHCP=no
IPADDR=10.22.22.22
NETMASK=255.255.255.0
DHCP_KEEPRESOLV=no

 

/boot/config/go

ifconfig eth1 192.168.1.116 netmask 255.255.255.0 up
route add default gw 192.168.1.1
# Start the unRAID Management Utility
/usr/local/sbin/emhttp &
# ...

 

Plug the network cable into your new NIC, and when the server boots, point your web browser to 192.168.1.116

 

...If you have no way to disable the onboard NIC in BIOS...

 

What's the old saying: When all else fails read the manual.  The Super MB has a jumper (strangely the idea of checking for a jumper switch came to me while dreaming last night) to switch off the on board LAN (hardware not software). I moved the jumper and it's now it's working off the Ethernet card.  After spending over an hour yesterday and walking up/down stairs a dozen times it took less thaN 5 minutes today.  Now I just have to wait and see if that was the problem.  If this thread dies then I'm happy and problem fixed.

 

Thanks

let us know how it goes, i've been having something similar, i was copying a load of stuff to a disk using teracopy and doing crc checks afterwards, when i checked it this morning they had all stalled, and was having some issues connecting to the disk shares. I tried ifconfig eth0 down && ifconfig eth0 up from ssh but when it came back up it was only at 100mbps, so i logged in locally and tried it, same thing, pulled the ethernet out of the back and the light stayed on on the ethernet port on the back panel, which was weird, plugged it back in, this time it came up 1gps, but still couldn't access the disk share, then a couple of mins later i could.

Not really sure if its an issue with the server ethernet port (which i blieve is an nvidia ethernet port) or with the htpc's ethernet port, or even the cables.  But i have a couple of intel pro adapters lying around at my parents, so will pick one up tonight and see how i get on with it.  I could also try disabling the nvidia mac lan and see if eth1 is any better.

i have a couple of intel pro adapters lying around

intel pro are proven rock solid with unraid.

 

I could also try disabling the nvidia mac lan and see if eth1 is any better.

when you disable the onboard nic in bios, the new nic will show as eth0.

 

i have a couple of intel pro adapters lying around

intel pro are proven rock solid with unraid.

 

I could also try disabling the nvidia mac lan and see if eth1 is any better.

when you disable the onboard nic in bios, the new nic will show as eth0.

 

What i meant is that there are 2 onboard nics, according to the manual

- PCI LAN supports 10/100/1000 Fast Ethernet by Marvell 88E1115 PHY.

- PCIE LAN supports 10/100/1000 Fast Ethernet by Marvell 88E8053.

 

Although the BIOS has

- nVidia MAC LAN

- OnBoard LAN Controller

 

So i have no idea which is which, but taking a wild stab, the top one is probably the nVidia MAC LAN (which i think also has a hardware firewall which may cause data corruption according to some nforce4 forum reports)

 

But i will probably disable both and put an intel pro adapter in, as i have them lying around doing nothing anyway.

Just to report: 8 days and not a single drop. While probably not a record it's not often it would go a full week without at least 1 drop.  I had to power down because I'm adding 2 drives.  The 1st is on preclear now and after the 2ed is through I will need to put 1 in the parity since it a larger size.  I will need to power down again because I want to do some swapping around (will need to read the Wiki and relearn) so won't get a fresh start until Sunday or Monday.  I will report back if I make it a month without a drop.

 

Also besides going over a week without a drop another problem I was having has disappeared.  When I would start a movie from a stopped HD in the first 2 minutes if would get near freezes and slow video that would get out of sync of the audio.  the video would then catch up.  I would always let it play with sound muted for a couple thinking the disk needed time to properly spin up.  Now the movies play perfect from a just started HD.  Looks like the on board LAN board was indeed bad. I'm feeling optimistic about the new card. ;D