ASUS M4A78L-M passes unRAID level 1 test


Recommended Posts

ASUS M4A78L-M uATX motherboard passes unRAID level 1 test.  See attached syslog.  I have a 3-drive unRAID system.

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16813131608

 North Bridge AMD 760G

 South Bridge AMD SB710

 SATA 3Gb/s 6 ports

 LAN Chipset Realtek 8112L

 Max LAN Speed 10/100/1000Mbps

AMD Sempron 140 Sargas

2x1GB Kingston ddr2800 RAM

Lexar 2GB Firefly flash drive

 

I pulled out a gigabyte MB and installed this Asus M4A78L-M on an existing unRAID system.  It Booted unRAID correctly first time, found all 3 hard drives (2 data, 1 parity) but they were on different sata ports so I went to the devices page and corrected them.  Each drive has a green  ball after that.   I rebooted and started the array.  I selectd the button for parity check.  The parity check completed with zero errors.

 

The MB bios detected the flash drive and allowed it to be chosen as the primary boot device.  This board supports the 'gracefully shutdown' when using the power button to turn it off.  S3 isn't working for me on this Asus board.  It did work on the Gigabyte board I pulled.

 

The board is marked Rev. 1.02G and it came with bios 0401.  I downloaded and flashed bios 0502.  Note: It unzips as M4A78L-M-0502.ROM but you have to rename it 0502.ROM to do the flashing because if the name is longer then there is an error and flashing doesn't happen.  To flash the board from the bios page:  copy the 0502.ROM to flash drive...boot machine and hold DEL key to enter bios settings...Tools/EZ Flash 2...Select 0502.ROM.  

 

 

Additional:

 

I liked the layout on the GA-MA74GM-S2 much better than this Asus board.  On the Gigabyte goard the PSU connector is higher up out of the way and the placement of the cpu and ram slots were about an inch further towards the io shield side leaving a little more breathing space for the hard drives to hang over the MB.  However, my RAM is running much much cooler on the Asus than it did on the Gigabyte so I don't have to use the RAM cooler fan so maybe it's a wash.  I'm replacing the Gigabyte board to get 6 sata ports instead of just 4.

syslog-2010-03-13.zip

Link to comment
  • 1 month later...

Just a quick update on the motherboard.  I found a fix to an annoying problem.  Every time I added or removed a sata or esata drive and booted the bios would change the order of the drives and the flash drive would move down the list - so it wouldn't boot from it. 

But I just discovered that in bios you can select any drive and tell bios to treat it as a (hard drive, floppy, CDRom).  Setting the flash drive to FORCE FDD then it will boot from it every time no matter how the other drives are rearranged.

 

 

Link to comment
  • 2 weeks later...

But I just discovered that in bios you can select any drive and tell bios to treat it as a (hard drive, floppy, CDRom).  Setting the flash drive to FORCE FDD then it will boot from it every time no matter how the other drives are rearranged.

 

Good to know... I'm building my server with the same board and CPU...

 

Link to comment
  • 1 month later...

I want to report that I just rebuilt my unRAID sever today with total of 11 hard drives using this motherboard. Six drives on the onboard SATA, and five on the PCI-SATA controller (forgot the exact model). I'm still in the process of parity check, but so far everything looks and works fine.

 

All I tweaked in the BIOS were:

 

1. Set the USB device as "forced FDD"

2. Changed the boot sequence to USB first, and disabled 2nd and 3rd options.

3. Made sure all 11 hard drives were listed in BIOS.

 

I also had to upgrade from 4.3.3 to current 4.5.4 since 4.3.3 didn't recognize the motherboard devices properly. I was asked to assign the hard drives to the array at first starting of 4.5.4. Luckily I printed the main page of server few months ago to a PDF file and emailed it to myself. That was of immense helps...

 

It's kind of nerve-racking to replace the motherboard in a server of 11 hardrives (6x2TB, 3x1.5TB and 2x1TB, and all are nearly full), but I'm very happy that despite several small glitches all went through rather smoothly.

 

The server had been running with ASUS P5B-E for over three years without major glitches. But weeks ago the SATA 0-5 ports on P5B-E all died, forcing me to change the motherboard. I decided on the M4A78L b/c I am scared by the HPA issue on the Gigabyte boards.

 

 

 

 

Link to comment
  • 4 weeks later...

I just built my first unRAID server last week using the ASUS M4A78L-M mobo and so far everything seems to be working fine.  I have 4 hard drives in my system and I'm sure I'll add more later once my storage needs grow.

 

The hard drives go to sleep as expected, wake on LAN works, when I push the power button on the computer case the system performs a graceful shutdown.

 

I also observed the BIOS flash drive boot order shuffling issue but I think I've isolated this behavior to whenever a new hard drive is added to the array.  I will try the "force FDD" BIOS tweak the next time I have a monitor and keyboard hooked up to it. 

 

So far no complaints!

Link to comment

UPDATE: I spent some time working on putting my unRAID server to sleep when it's not in use (using the ASUS M4A78L-M mobo) and everything seems to be working fine.  The server goes to S3 sleep after x number of minutes of hard drive inactivity and it re-awakens whenever it receives the "magic packet" from any PC on our home network.

 

I accomplished this by following the instructions found here:

 

http://www.lime-technology.com/wiki/index.php?title=Setup_Sleep_%28S3%29_and_Wake_on_Lan_%28WOL%29

Link to comment
  • 1 month later...

To further elaborate, when I have two drives in the system, I can select the USB Flash drive as the boot device, and boot to unRaid, no problem. 

 

As soon as I add the a drive, I can no longer select the Flash drive in the dropdown (in the Boot Hard Disk Drive menu).  Only the three HD's come up - the USB selection has disappeared.  I'm stuck, unless someone can help me find this "forced FDD" setting (nothing under USB config menu or anywhere else that I see).

Link to comment
  • 2 weeks later...

Hi everyone,

 

I recently build a budget unRaid box using the Asus motherboard and while everything else if working fine I'm having some issues in getting the server to resume from S3 sleep. I've been trying to use the S3 script that is elsewhere on this forum but removed it to try and isolate the problem. I've updated the BIOS to the latest one and enabled S3 only, PME - Enabled, ACPI - Enabled and APIC - Enabled. Now here come the strange part, if I put the server to sleep via the user scripts page I can wake it up by sending a magic packet within a few seconds of the drives spinning down, but wait anymore than that e.g. 30 minutes and sending the magic packet does nothing at all. The only way I can wake up the server is by manually pressing the power button.

 

So has anyone had any success in getting this working, and if so what settings did you set in the BIOS in case I've missed something.

Link to comment

I recently build a budget unRaid box using the Asus motherboard...

 

Just to clarify, are you using the ASUS M4A78L-M motherboard?

 

So has anyone had any success in getting this working, and if so what settings did you set in the BIOS in case I've missed something.

 

If you are using the M4A78L-M motherboard then I'll check my BIOS settings and let you know once I return home tomorrow... unless someone else beats me to it.

 

 

Link to comment

Here's my BIOS settings.

 

BIOS version = 0705

 

Main

SATA Config --> OnChip SATA = ENABLED

SATA Config --> SATA port1 - port4 = AHCI

SATA Config --> SATA port5 - port6 = AHCI

 

Advanced

Onboard Devices Configuration --> Serial ports = DISABLED

Onboard Devices Configuration --> Parallel ports = DISABLED

Onboard Devices Configuration --> Audio = DISABLED

Onboard Devices Configuration --> LAN controller = ENABLED

Onboard Devices Configuration --> LAN boot ROM = DISABLED

 

Power

Suspend mode = S3 Only

ACPI 2.0 support = ENABLED

ACPI APIC support = ENABLED

 

APM Config --> (all at default settings)

Hardware Monitor  --> (all at default settings)

 

Let me know if you have any luck with the above.

 

Link to comment

I've replicated your settings and so far the server won't wake up, I've looked at the status lights on the NIC and they're both out which is strange...

That's how it acts on my newer server.  To me it indicates the network adapter does not have power,and therefore is not listening for any wakeup.

 

I cannot wake it either by a WOL packet... but then I don't usually put it to sleep. (other than when I wrote that "button" for unMENU)

Link to comment

I've replicated your settings and so far the server won't wake up, I've looked at the status lights on the NIC and they're both out which is strange...

 

I tested putting my server to sleep using the "Go to S3 Sleep" button on unMENU and the WOL/magic-packet works fine for me both immediately after I put the server to sleep and 60+ minutes after I put the server to sleep.

 

I wonder if your server does a hard powerdown instead of going to S3 sleep as you are expecting. 

 

Is there anything in your syslog file that may give some clues?

 

What WOL / magic-packet tool are you using?  What syntax are you using to send the magic-packet to your unRAID server?  Are you sure the MAC-address is correct?

 

My WOL command looks like this:  C:\utils\wolcmd  485b9548ed03  255.255.255.255  255.255.255.255

 

What is your network type?  Are you doing this over a wired or wireless network?  Have you tried issuing the WOL command from a different PC on the same network?

 

Also, are you using the NIC that came integrated with the ASUS mobo?

Link to comment

Hi Guys,

 

to follow on from my last post I've tried out a couple of things, according to one of the threads on this forum relating to s3 and WOL there was some mention of running Windows and making some changes to the device's power management settings, I tried using a Hiren's boot disc and running Mini XP and tried altering the power management settings for the onboard NIC which strangely was being detected as a Realtek 8168/8111. Tried rebooting unRaid and got the same results.

 

Here's the contents of my wakeup batch file:

 

cd c:\wolcmd

wolcmd.exe 485B39622053 192.168.1.100 255.255.255.255

 

I've tried both sets of 255's with no change.

 

My network is wired, I have a 10/100 switch and a gigabit switch I've tried running the script from my HTPC with the same effect.

 

 

Now this might be of interest. In the syslog the onboard NIC is detected as an r8168 but it loads the 8169 driver...

 

Sep  7 21:30:14 TARDIS kernel: r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded

Sep  7 21:30:14 TARDIS kernel: r8169 0000:02:00.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18

Sep  7 21:30:14 TARDIS kernel: r8169 0000:02:00.0: setting latency timer to 64

Sep  7 21:30:14 TARDIS kernel: r8169 0000:02:00.0: irq 26 for MSI/MSI-X

Sep  7 21:30:14 TARDIS kernel: eth0: RTL8168d/8111d at 0xf8276000, 48:5b:39:62:20:53, XID 083000c0 IRQ 26

 

Now according to this site http://louwrentius.blogspot.com/2009/12/problem-with-wake-on-lan-wol-and-r8168.html there's an issue with WOL not working due to the driver and the proposed solution was to backlist the 'r8169' driver in the blacklist.conf file.

 

Unfortunately I've not been able to get the change to be saved back to the flash drive to run on the next boot.

 

There's one more weird thing... if I open a telnet session and login as root, if I enter the command "echo 3>/proc/acpi/sleep" the server doesn't go to sleep but runs a parity check  ???

syslog-2010-09-07.zip

Link to comment

My syslog contains the same entries as yours (regarding the NIC) yet S3 / WOL is working just fine for me:

 

Sep 6 11:51:46 Titan kernel: r8169 Gigabit Ethernet driver 2.3LK-NAPI loaded
Sep 6 11:51:46 Titan kernel: r8169 0000:02:00.0: PCI INT A -> GSI 18 (level, low) -> IRQ 18
Sep 6 11:51:46 Titan kernel: r8169 0000:02:00.0: setting latency timer to 64
Sep 6 11:51:46 Titan kernel: r8169 0000:02:00.0: irq 26 for MSI/MSI-X
Sep 6 11:51:46 Titan kernel: eth0: RTL8168d/8111d at 0xf8276000, 48:5b:95:48:ed:03, XID 083000c0 IRQ 26

 

Some other random thoughts:

 

I have a static IP-address assigned to my unRAID server.  My client PCs all use DHCP.  If you are using DHCP with unRAID, maybe test it with a static IP ?

 

Try plugging your unRAID server into the same network switch that your client PC is using.  If that doesn't work, try plugging both into the other switch and test that.  I'm wondering if there's possibly some issue with using more than one network switch.

 

 

Link to comment

Hi there,

 

well my unRaid box has a static IP and I tried running the wakeup script from my HTPC which is on the same switch as the unRaid box with no change. It looks like the server is going into a deep sleep and powering off the onboard NIC. I'm going to try and dig a little deeper into the driver issue as I am sure there must be some subtle differences between the 8169 and 8111 drivers.

 

To be continued...

Link to comment

And now the thrilling conclusion to this episode  :o

 

This has been bugging me for days now, it's a good job I lost my hair years ago else I'd be bald now  :P

 

I started wondering if I had a faulty board so I decided to fit an Intel Pro 100 NIC and disable the onboard one and try testing again. I noticed that a new MAC address was being picked up during the boot phase and made a note of it and tried testing the WOL function again. The result was still the same... nothing... While reading about WOL and how it works I found a nice Windows based magic packet sender from http://magicpacket.free.fr/.

 

While using the program I suddenly noticed something that I should have spotted earlier... you know the kind of problem where the answer is staring you in the face and you go off in every other direction because it couldn't possibly be that simple... Well the old saying "fool me once... shame on you... fool me twice... shame on me!" came true. The problem was due to the subnet mask! The value 255.255.255.255 is a broadcast to all networks (like the internet!) which would take who nows how long before getting to my internal network. As soon as I changed the subnet mask to 255.255.255.0 (which is right for a 196.168.0.0 IP address) the server sprang to life. I've tested it several times from my PC which is linked through another switch and it's woken up every time like a champ!

 

For future reference the "correct" format of a Wake On Lan command should be:

 

wolcmd.exe {Server MAC Address} {Server IP Address} {Server Subnet Mask}

 

eg: wolcmd.exe 485B39622053 192.168.1.100 255.255.255.0

 

Well at least it looks like this mystery is now solved.

 

Thanks to everyone who helped! It proves that even IT Admins can get it wrong  :-[

Link to comment
  • 2 months later...

going for this board, or the me4a785t-m

whats your opinion guys?

whats the difference between these 2?

 

any update on the next level for this board?

 

Hi - guess who? As far as I know, the "T" uses AM3 socket and the "L" uses AM2+. Also the "T" uses DDR3 memory and the "L" uses DDR2 (I thik I got that right).

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.