BUILD Building a server for FreeNas primary for Plex

[depasseg]

That's weird. I pasted the exact link (in fact the text is correct). I think the forum SW stripped it.

 

[AltecBX]

Success updating the 3008 to IT mode.
View attachment 5734

Steps:
0. Find the SAS ID (I found mine on a sticker on the motherboard (starts with SAS: 50003048) - write down the last nine characters.
1. Download the 3008_FW_PH6_091714.zip file and extract IT folder to a USB drive
2. Insert USB drive into server and power on
3. Press F11 to choose boot menu
4. Boot into EFI Shell
5. change to the USB device (type 'fs0: or type 'map' for a listing)
6. cd to IT/UEFI folder
7. run 'SMC3000T'
8. when prompted, enter the 9 characters from step 0. It's a quick timeout, and I finally tried lowercase hex which worked (the first couple attempts failed).

It's not working for me. I get to the command line and I enter in "smc3008t" and I get is "Bad command of filename"

That's weird. I pasted the exact link (in fact the text is correct). I think the forum SW stripped it.

That's what I thought happened too.

 

[depasseg]

tsk, tsk. you aren't in the EFI shell. You are in DOS. Either choose the EFI shell, and use the .nsh file, or copy the files from the DOS folder to your USB and try the .bat file.

And you might want to use the PH5 firmware, since I think that is what the FreeBSD is using. I'm going to try to downgrade.

I see this in dmesg:

Code:

[root@freenas] ~# dmesg | grep mp
mpr0: <LSI SAS3008> port 0xe000-0xe0ff mem 0xfb200000-0xfb20ffff irq 26 at device 0.0 on pci1
mpr0: IOCFacts  :
mpr0: Firmware: 06.00.00.00, Driver: 05.255.05.00-fbsd
mpr0: IOCCapabilities: 7a85c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,EventReplay,MSIXIndex,HostDisc>

 

[mjws00]

Was wondering when you'd check that. Also did you notice the memory install slots, depasseg?

 

[depasseg]

Slots are all full with identical sticks, so that should be OK.

Hopefully this is better:

Code:

[root@freenas] ~# dmesg | grep mp
mpr0: <LSI SAS3008> port 0xe000-0xe0ff mem 0xfb200000-0xfb20ffff irq 26 at device 0.0 on pci1
mpr0: IOCFacts  :
mpr0: Firmware: 05.00.00.00, Driver: 05.255.05.00-fbsd
mpr0: IOCCapabilities: 7a85c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,EventReplay,MSIXIndex,HostDisc>

PS - This is how my MB came cabled to the SuperMicro backplane. The white wires go from the MB to the Backplane, and the blue ones go from the backplane to 2 external ports on the rear for expansion.

 

[mjws00]

Heh. Filling all the slots takes the fun out of it. ;) Of course the sas3 backplane does as well. Thanks though. I do appreciate a sexy pic.

 

[depasseg]

If I didn't fully populate, SM sent me a handy guide for installing RAM: http://www.supermicro.com/QuickRefs/superserver/2U/QRG-1607.pdf

I know, takes all the fun out of it.

 

[mjws00]

Interesting that the chart matches exactly what the tech told AltecBX. Pretty much slam em in from left to right. Interesting change even from the x10sra board. They need to stick that page in the mobo manual.

 

[Ericloewe]

If I didn't fully populate, SM sent me a handy guide for installing RAM: http://www.supermicro.com/QuickRefs/superserver/2U/QRG-1607.pdf

I know, takes all the fun out of it. :)
View attachment 5737

I wonder who wrote that...

I'm not going to say they're wrong, but it's very weird, since it restricts operation to dual-channel mode instead of quad-channel mode. The channels can't possibly go to both sides, since that would make for an insanely complicated memory trace layout. So, that leaves the question of whether there's something weird going on.

If anyone is using these as a tinkering project (yeah, right, that's likely at these prices...) and is interested in doing science, I would appreciate feedback on the "traditional" DIMM placement of "fill all channels first, then go back, rinse and repeat".

 

[depasseg]

I'm still testing my setup before going to production, so if you can let me know what tests you'd like to see on what config, I'll run them and post the results.

 

[Ericloewe]

I'm still testing my setup before going to production, so if you can let me know what tests you'd like to see on what config, I'll run them and post the results.

Doesn't really apply to your case, since all slots are populated, but thanks for the offer ;)

 

[depasseg]

I'm willing to take some out for testing. lmk if you are interested. I still have to run my memory tests, so this could be a good stress test.

 

[Ericloewe]

I'm willing to take some out for testing. lmk if you are interested. I still have to run my memory tests, so this could be a good stress test. :)

In that case, if you don't mind, it'd be nice to know if any weird problems arise when slots A1, B1, C1 and D1 are populated, with the rest empty, and vice-versa.

 

[AltecBX]

tsk, tsk. you aren't in the EFI shell. You are in DOS. Either choose the EFI shell, and use the .nsh file, or copy the files from the DOS folder to your USB and try the .bat file.

And you might want to use the PH5 firmware, since I think that is what the FreeBSD is using. I'm going to try to downgrade.

I see this in dmesg:

Code:

[root@freenas] ~# dmesg | grep mp
mpr0: <LSI SAS3008> port 0xe000-0xe0ff mem 0xfb200000-0xfb20ffff irq 26 at device 0.0 on pci1
mpr0: IOCFacts  :
mpr0: Firmware: 06.00.00.00, Driver: 05.255.05.00-fbsd
mpr0: IOCCapabilities: 7a85c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,EventReplay,MSIXIndex,HostDisc>

I've tried everything from just putting those files on an empty USB drive to formatting a drive with Rufus to make it bootable. I used EFI and the Dos comand and nothing. I can't seem to get this to upgrade.

:)

Slots are all full with identical sticks, so that should be OK.

Hopefully this is better:

Code:

[root@freenas] ~# dmesg | grep mp
mpr0: <LSI SAS3008> port 0xe000-0xe0ff mem 0xfb200000-0xfb20ffff irq 26 at device 0.0 on pci1
mpr0: IOCFacts  :
mpr0: Firmware: 05.00.00.00, Driver: 05.255.05.00-fbsd
mpr0: IOCCapabilities: 7a85c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,EventReplay,MSIXIndex,HostDisc>

PS - This is how my MB came cabled to the SuperMicro backplane. The white wires go from the MB to the Backplane, and the blue ones go from the backplane to 2 external ports on the rear for expansion.
View attachment 5736

Wow. That's a beauty.

If I didn't fully populate, SM sent me a handy guide for installing RAM: http://www.supermicro.com/QuickRefs/superserver/2U/QRG-1607.pdf

I know, takes all the fun out of it. :)
View attachment 5737

So the Tech was right. Just leave my (4) 16GB sticks in A1. A2, B1,B2

Interesting that the chart matches exactly what the tech told AltecBX. Pretty much slam em in from left to right. Interesting change even from the x10sra board. They need to stick that page in the mobo manual.

I guess I'll leave it the way it is.

 

[AltecBX]

Depasseq, how did the PH6 differ from the PH5?
Is FreeNAS working to use PH6?
What's the difference in both drivers?

 

[AltecBX]

Finally got it flashed to IT mode MPT 5. :)

But unfortunately I only still see 9 drives in FreeNAS.

 

[depasseg]

Depasseq, how did the PH6 differ from the PH5?
Is FreeNAS working to use PH6?
What's the difference in both drivers?

I didn't notice any difference between PH5 and 6 (nor IT and IR for that matter).
The driver in FreeNAS looks like v5 and everything I read highly suggested keeping the driver and firmware versions the same.

what does 'dmesg | grep mpr' report?

 

[Mlovelace]

I wonder who wrote that...

I'm not going to say they're wrong, but it's very weird, since it restricts operation to dual-channel mode instead of quad-channel mode. The channels can't possibly go to both sides, since that would make for an insanely complicated memory trace layout. So, that leaves the question of whether there's something weird going on.

If anyone is using these as a tinkering project (yeah, right, that's likely at these prices...) and is interested in doing science, I would appreciate feedback on the "traditional" DIMM placement of "fill all channels first, then go back, rinse and repeat".

I agree, it's curious that the interleaving would use two different patterns between the X10SRA and X10SRH 2011R3 boards, and also depart from all the previous socket 2011 interleaving patterns. It doesn't really make sense to me, and it looks as if the manual for the X10SRH board is incomplete. It states " Populating these DIMM modules with a pair of memory modules of the same type and same size will result in interleaved memory, which will improve memory performance. Please refer to the table below:" but there is no table below.

 

[depasseg]

There's a table above. Post #127.

 

[AltecBX]

I didn't notice any difference between PH5 and 6 (nor IT and IR for that matter).
The driver in FreeNAS looks like v5 and everything I read highly suggested keeping the driver and firmware versions the same.

what does 'dmesg | grep mpr' report?

I'm lost with that one. How can I show you?

Post the output of 'dmesg' in code tags. You could also hit the debug button and post the file.

If I were troubleshooting I'd isolate the controller and backplane and validate proper function. For me that would be windows, Linux, or bsd. I'd also skip the beta until things were proven. Could be a gui glitch or something in the freenas driver, but could also be expander controller related. The bios showing the drives is great but only step one.

Here's Ubontu showing 10 drives.