Supermicro trays shines red (1U, on board SATA)

Yarik Dot

Active Member
Apr 13, 2015
213
109
43
46
Hi,

I got a problem, which I don't know what causes it. I have 1U supermicro server with 4 SATA ports connected directly to the on-board.

The backplane still let HDD trays to shine red (no blinking). Seems like I miss a jumper somewhere, because we replaces MB and it still persists.

No matter how many drives I populate. It's still on for all ports.

Any idea how to turn it off?

Thanks

WP_20180212_22_39_28_Pro.jpg
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,125
575
113
New York City
www.glaver.org
I got a problem, which I don't know what causes it. I have 1U supermicro server with 4 SATA ports connected directly to the on-board.
How are these connected? 4 individual SATA cables? Multi-lane cables? Something else?

It would also help to know both the motherboard model and the chassis model (exact model numbers, please, not generic).
The backplane still let HDD trays to shine red (no blinking). Seems like I miss a jumper somewhere, because we replaces MB and it still persists.
This sounds like an SGPIO / iPass / sideband (all names for the same thing) issue - the red LEDs are normally controlled by a SAS controller. Using (for example) an SFF-8087 cable without sidebands can give you this sort of behavior. So could an improperly configured backplane. Compare the systems that work normally with the one that does not, paying special attention to the controller-to-backplane cable(s) as well as the backplane jumpers (and possibly model / revision if different between working and non-working chassis).
 

svtkobra7

Active Member
Jan 2, 2017
362
84
28
This sounds like an SGPIO / iPass / sideband (all names for the same thing) issue - the red LEDs are normally controlled by a SAS controller. Using (for example) an SFF-8087 cable without sidebands can give you this sort of behavior. So could an improperly configured backplane. Compare the systems that work normally with the one that does not, paying special attention to the controller-to-backplane cable(s) as well as the backplane jumpers (and possibly model / revision if different between working and non-working chassis).
I was thinking this was SGPIO related (but figured I'd let an expert chime in). I have an SC836 with 2.5" HDD Hot-Swap Kits installed where one might otherwise find a DVD / FDD drive:
Not having connecting the SGPIO cables, IPMI View shows "Rebuild/Remap in progress" for those guys, but I don't get a red light (blue on access though). 1 of the Hot-Swap Kits is connected via SATA directly to the motherboard (X9DRi-LN4F+), the other reserved for future use.



@Terry Kennedy: Any advantage to connecting the SGPIO cable (other than ridding myself of an error message in IPMI View)? Also, where do you connect the SGPIO cable(s) on that board (not sure I follow the bit about connecting to Port #0, or how that is achievable with two kits)?

 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,125
575
113
New York City
www.glaver.org
Not having connecting the SGPIO cables, IPMI View shows "Rebuild/Remap in progress" for those guys, but I don't get a red light (blue on access though). 1 of the Hot-Swap Kits is connected via SATA directly to the motherboard (X9DRi-LN4F+), the other reserved for future use.

I think that's a different issue. I'm not sure where IPMIview gets that data from, but it is unlikely it is trying to read it from the 2.5" hot swap bay. I'd suggest contacting Supermicro support to see what they say (and to clarify the manual section you quoted).
 

svtkobra7

Active Member
Jan 2, 2017
362
84
28
I think that's a different issue. I'm not sure where IPMIview gets that data from, but it is unlikely it is trying to read it from the 2.5" hot swap bay. I'd suggest contacting Supermicro support to see what they say (and to clarify the manual section you quoted).
There aren't many times I'd disagree with you, but on this one I have to, as it appeared immediately after adding the two bays and there are two sensors. Nevertheless, I will take your advice! Thanks for weighing in and hope all is well.
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,125
575
113
New York City
www.glaver.org
There aren't many times I'd disagree with you, but on this one I have to, as it appeared immediately after adding the two bays and there are two sensors.
The manual (here) says the backplane only has 3 connectors - power, SATA and GPIO. The fault signal isn't on the power or SATA connectors, so if you don't have the SGPIO cable connected, I don't see how IPMIview could read anything from the backplane.
 

Yarik Dot

Active Member
Apr 13, 2015
213
109
43
46
I will check the jumper setting when there is planned outage window for that server. Hopefully it should be by the end of this month.
 

james23

Active Member
Nov 18, 2014
432
104
43
50
did this get resolved? im seeing the same thing on a bpn-sas-846TQ but only on the bottom most 2 rows. at one point (in my 1-2 months of testing an OS on this system) the red leds did go off, but ofcouse came back.

even with disks in all the bays, the red leds on the bottom two rows stay on (but the blue leds flicker as expected). Ive checked all the fan jumpers (its 2x jumpers per fan header, so 8x total), and i have them all set to 2-3 (which is NO FAN, as i have all fans connected to my MB, an x9dr3-ln4f+) i dont have any sgpio cables connected to the BP (should this go from the raid card to the BP, correct?)

i have 4x devices connected to the BP via fan out cables (a lsi 9207-8i , and 9207-8e , 4 of the 6 MB sata ports, and 2x of the MB intel SCU 8087 ports) for a total of 24 connected sata plugs.

ive read every page of the SM BP manual, and cant find anything related to this.

(i dont have that 0% raid rebuilding thing showing up in IPMIview, on this machine. but i have seen that on other machines before)

any ideas? tks
 

james23

Active Member
Nov 18, 2014
432
104
43
50
EDIT/UPDATE: well crap, they are back on red again, (after some reboots and changing some HW around for testing of something unrelated, its the same bottom two rows, horizontally, as before. however, i am on the right track tho, and it is related to the jumpers below, i just need to work on it a bit more and will update. It has something to do with the jumpers below + that i have 4x devices connected to this TQ 24bay BP- 2x LSI 9207 cards, the onboard intel SCU w 2x 8087 ports, and a few sata ports from the intel MB sata ports). im on the right track though and will be able to get the lights off permanently . will update

----
i figured it out!
this post from the SM FAQs - directs a user with similar issues (2 rows of all SOLID red LEDs, on a 24x backplane), to check that he has his SGPIO / i2c jumper (on the BP) properly set.
FAQ Entry | Online Support | Support - Super Micro Computer, Inc.

On my 24 bay TQ BP, there are 2x of these jumpers. one was set to 1-2 (SGPIO) the other was set to 2-3 (i2c). Im not sure which jumper controls which set of rows, but through trial and error (must restart each time), i figured out i needed to have the JUMPER OFF/REMOVED on both to get my LEDs. the jumpers im refering to are JP84 and JP189
(and now the 8x red leds are off!)

here is the PDF manual for my BP (bpn-sas-846tq) :
https://www.supermicro.com/manuals/other/BPN-SAS-846TQ.pdf
 
Last edited:
  • Like
Reactions: bicilis

bicilis

New Member
Mar 21, 2021
1
0
1
I have the same problem but with the two central rows. Where is JP189? I can't find it in the manual.