Supermicro backplane and mysterious HDD activity leds

ponky

Limp Gawd
Joined
Nov 27, 2012
Messages
178
I have been using a Supermicro M28SAB 8in2 2,5" hot swap backplane for a year now. There is one thing that has been bothering me from the start; with some SSDs the activity leds work in reverse. They stay on all times and go off when there is disk activity. HDDs and Intel X25M-G2 SSDs seem to work fine, but ALL Sandforce based and Samsung drives have this issue. I have checked the manual for possible jumper changes but no luck there.

I contacted Supermicro support about this. Here's the response:

"The activities LED is control by hard drive on pin 11. By the default, if user using SAS hard drive the LED will turn on solid and blink when there is activities. If using SATA hard drive the LED will be turn off and blink when there is activities."

I don't think that all Sandforce and Samsung SSDs are SAS....

Has anyone else experienced this or is there a logical explanation? I know it's not a world breaking issue but feels kinda abnormal, especially on a backplane at this price.
 
I have seen this from time to time with hard drives at rare occasions on their 3.5 inch models.
 
It seems that for SATA the LED is controlled by the drive itself while for SAS it is controlled by the controller. With some older backplanes (chenbro) I could choose whether the LED reacts to the drive or to the SGPIO signal by jumpers, the newer backplanes do no longer have these jumpers.

But my experience with SAS is very limited.
 
It still does not make any sense. Why do drives with Intel controller work and rest consumer drives dont? It feels like they just ignored the 11th pin on most SSD controllers, or maybe it's something deeper.
 
Samsung has had this problem (or a related problem where no activity light control is provided) on many versions of their SSDs, not just the Sandforce ones.
 
Hmm. There are only 7 pins on an SATA cable..
esata.gif


I think the enclosure actually snoops the cable for activity and when activity happens it toggles the led. Although with a fast drive it may not be able to keep up.
 
It seems that for SATA the LED is controlled by the drive itself while for SAS it is controlled by the controller. With some older backplanes (chenbro) I could choose whether the LED reacts to the drive or to the SGPIO signal by jumpers, the newer backplanes do no longer have these jumpers.

But my experience with SAS is very limited.

I've tried with a sideband cable (sgpio), a cable without sideband and even with no cable connected, led behaviour is always same (ofc there can't be activity when cable is not plugged, but sandforce / samsung drives still keep the led on solid).
 
They mean Pin 11 on the SATA POWER connector, not the DATA connector.

Pin 11 is an interesting pin. It is commonly used to manage staggered spin-up of drives in backplanes. (Not all though, some drives support it by commands only, and some don't support it at all).

If pin 11 is high or floating (as provided by most backplanes, but NOT standard sata power connectors) the drives will do staggered spin-up controlled by backplane.

If pin 11 is grounded then it'll spin up immediately.

Now, after spin-up, pin 11 is used to indicate activity LED's for most drives. Again, this wont work unless the backplane supports it.
 
They mean Pin 11 on the SATA POWER connector, not the DATA connector.

Pin 11 is an interesting pin. It is commonly used to manage staggered spin-up of drives in backplanes. (Not all though, some drives support it by commands only, and some don't support it at all).

If pin 11 is high or floating (as provided by most backplanes, but NOT standard sata power connectors) the drives will do staggered spin-up controlled by backplane.

If pin 11 is grounded then it'll spin up immediately.

Now, after spin-up, pin 11 is used to indicate activity LED's for most drives. Again, this wont work unless the backplane supports it.

Really nice explanation. It is still unclear why some drives work and some don't. It's pretty obvious that this backplane supports indicating activity by LEDs.
 
There are a lot of "vendor specific" comments in the SATA specification with regards to activity lights and staggered spin-up. If I'm reading it correctly it can either support staggered spin-up OR activity indication, however that doesn't make much sense and I doubt anyone follows that recommendation.
Anyway, pin 11 is supposed to be pulled low by the DRIVE to indicate activity, or "outstanding signals". There's also this quote "Devices may omit asserting the activity signal for commands that do not access the media and have an expected service time too short to allow visual perception of the signal."

Reading the SAS documentation there is this note:
"SATA devices use the pin used by the READY LED signal (i.e., P11) for activity indication and staggered spin-up disable (see SATA). The output characteristics differ from those in table 68"

Also, "A SAS target device uses the READY LED signal to activate an externally visible LED that indicates the state of readiness and activity of the SAS target device."
With lots of references to SPL-2... too much PDF junk for me to dig through now, sorry :)


So, my somewhat naive interpretation of all that is that SAS drives keep the LED lit to indicate READY, and when they're busy.. well, the LED is off.
SATA has the exact opposite behavior.

I would ask the manufacturer of the drives why their drives seem to be confirming to the SAS specification instead of the SATA one when it comes to activity signalling.
Of course, I may be entirely wrong about all of this as well :)
 
There are a lot of "vendor specific" comments in the SATA specification with regards to activity lights and staggered spin-up. If I'm reading it correctly it can either support staggered spin-up OR activity indication, however that doesn't make much sense and I doubt anyone follows that recommendation.
Anyway, pin 11 is supposed to be pulled low by the DRIVE to indicate activity, or "outstanding signals". There's also this quote "Devices may omit asserting the activity signal for commands that do not access the media and have an expected service time too short to allow visual perception of the signal."

Reading the SAS documentation there is this note:
"SATA devices use the pin used by the READY LED signal (i.e., P11) for activity indication and staggered spin-up disable (see SATA). The output characteristics differ from those in table 68"

Also, "A SAS target device uses the READY LED signal to activate an externally visible LED that indicates the state of readiness and activity of the SAS target device."
With lots of references to SPL-2... too much PDF junk for me to dig through now, sorry :)


So, my somewhat naive interpretation of all that is that SAS drives keep the LED lit to indicate READY, and when they're busy.. well, the LED is off.
SATA has the exact opposite behavior.

I would ask the manufacturer of the drives why their drives seem to be confirming to the SAS specification instead of the SATA one when it comes to activity signalling.
Of course, I may be entirely wrong about all of this as well :)

That would make sense. Can you link the PDF?
 
I have some 2Tb seagate SATA drives that do the LED always on and blinks off with activity in an old xServe G5, so it's not just supermicro stuff... never really investigated it further...
 
Back
Top