SOLVED Supermicro X10SL7-F and SAS Drives not found

[Mike Fuller]

Hi Jailer,

Sure, I have no qualms. They are just paperweights as far as I am concerned. Anyone with a known working SAS, reach out to me privately and I can mail one out.

 

[Mike Fuller]

Quick update...just browsing around in the BIOS...I notices the spin-up on boot is disabled in the SATA Controller by default, I am trying to CTRL+C into the LSI Legacy BIOS to see if I can set a similar option (probably useless since it can't see the drive at all) thinking maybe it will unlock the juice to the motor.

Also too...the LSI SAS Controller legacy BIOS seems to "disappear"...that is to say on a cold boot I don't see the option to get into it. I did a reflash to the IR BIOS and I had the option one one reboot, then it went away again. Not sure if anyone who has this board has ever seen this. I know when you flash to IT mode (which is popular here to make sure the controller RAID is OFF) you loose the legacy menu altogether. I know switching to UEFI instead of Legacy in the advanced settings moves the menus to the Mainboard BIOS menu - but the options there seem less than the legacy. Just trying to see if I can spot any additional error messages other than drives connect is zero :)

I did get the SATA Controller to properly populate my SSD again - but now for whatever reason its not seeing the boot sector. I'll double check the UEFI link, but it just seems like more odd behavior. I am beginning to think its my Supermicro Board that needs an RMA....

 

[Mike Fuller]

Haha...Well reflashing everything caused me to loose my freeNAS pointer in the UEFI...I am not sure how to get that back without reinstalling the OS. If I use the Driver Pointer to x64 in the BIOS, I can get it pointed back to the SSD but FreeNAS kernel panics before I get too far into things.

I appreciate everyone's help, but I know when I have spent enough time and money with zero results to know when its time to scrap a project and go back to what I know works. Pretty expensive learning experience, but I did conclude the following:

1. Stay away from SAS at all costs unless you are truly an enterprise geek and you know what you are doing. Besides, looking at price and performance - there is zero reason to buy SAS anything - especially given all the hardware issues you could possibly encounter. Most hobbyist folks on here seem to use plain jane SATA drives, I should know at this point in my life you never steer off the beaten path. Update 05/13/2017 even though I got these SAS drives to work, I still stand by this comment.

2. Seagate RMA service took weeks to round trip everything. Not sure if WD or Hitachi is any better, this was my first RMA experience with hardisks. Its probably better to have spare drives in the locker than to expect a fast turn around from a vendor if you need to do a disk replacement in you RAID / ZFS array.

3. Supermicro back panels for their serve boards are junk. Not sure why they elected to have razor sharp tabbies that do not break off with a little bending.

4. Flashing Supermicro Mainboard and LSI Controller BIOS is seemingly a lot easier in UEFI Shell than DOS Shell. Mounting drives is different from Linux.

5. Registered ECC RAM cannot be used with this Supermicro board. Strict attention should be paid to the RAM compatibility tables to ensure you get the right type (Unbuffered / ECC) at the right clockspeeds. This has been discussed in more detail in other threads - pay attention to it.

6. Supermicro tech support was very fast and at least tries to be very helpful Update 05/13/2017, they were right

7. Seagate tech support is non-existent Update 05/13/2017, it is existent, just slow

8. Folks on this forum definitely try to be helpful and I am glad my buddy who talked me into this silly project got me here. Its good to have a quick responding community that is there to support your endeavor.

9. My linux server has an uptime I am measuring in months...why I thought I *needed* to spend the time and money with FreeNAS is unclear to me. If its not broke, don't fix it. I think I am a glutton for learning new things.

10. What little I did see of FreeNAS itself seemed nice. The Web GUI was excellent.

I am going to box up the motherboard / processor / ram / drives and send them to my buddy in Boston so he can play with it if he feels so inclined. I hate to toss stuff in the garbage, even though smashing these things with my 30-06 would undoubtedly provide some small juvenile satisfaction which is about all I can get out of this with 800 bucks in the toilet. If its all the same to the folks here - I will have have him re-open this thread if he wants to try some more.

Thanks again,
-Mike

PS my offer still stands if someone (or two) people want to try one of these new drives I just got back from my Seagate RMA. If anything, I would feel a little better knowing we could vindicate Seagate and in fact put all the blame on this loser of a system builder (me)!

 

[Jailer]

Well I do thank you for giving it a good go. It's too bad we couldn't be more helpful with your build and I hope you don't let this experience sour your ecperience to FreeNAS, it is a truly wonderful storage solution.

 

[Spearfoot]

@Mike Fuller : I'm sorry for your troubles with this server build-out! Is it possible you don't have the onboard SAS controller enabled in the BIOS? Here are the relevant screenshots from my X10SL7-F BIOS setup (notice that I don't load the SAS controller option ROM, because I don't boot from it - your situation may require that this be enabled):

This is a long shot, but it's all I got...

 

[Mike Fuller]

All,

I found this post here:

https://serverfault.com/questions/8...y-recognises-sata-drives-but-refusing-to-reco

Seems like this fella had the exact same troubles I was having. I also finally did hear back from Seagate tech support who verified that all the RMA drives I received were all tested just prior to shipping and they blamed the cable or the controller. Supermicro said that while it COULD be an issue with their board or the controller, they tested these model drives and confirmed they work (albeit at 6GB/s instead of the 12Gb/s they are capable of). So based on advice here, and advice from the vendors, and reading that above post very carefully I decided, since I already have close to 1000 bucks invested in this FreeNAS box, I mind as well pony up another 18 bucks for a higher quality cable.

I purchased some Coboc SFF8482-SATA-0.5M cables. I have always had good luck with their displayport and HDMI cables, and they are reputable. In fact on Newegg, someone commented they had this exact board and a SAS drive they needed connecting and this worked (after also complaining several cables were junk).

So this morning, I hooked it up, and sure enough, there in the BIOS under the LSI Controller physical disk map are my Seagate SAS drives confirmed and functioning. FreeNAS works again since I re-flashed back to the IT BIOS, and its ready to start building my ZFS array.

So...lesson of my SAS story is...buy high quality cables from a reputable manufacturer. I tried two different bargain bin cables from Amazon, and while they worked good with SATA, they did not work with SAS. Something is wired differently someplace somehow. I may just buy an extra cable to dissect and see EXACTLY how one of these cables is wired different than the other.

Also, these particular Seagate drives do NOT spin up with power attached. Seagate tech support also confirmed this is true - the drive platter will only spin up when the controller is ready to access the platter for data transfer. So if anyone else is troubleshooting, you should be able to see the drives over the data cables.

Thanks to everyone for their patience and suggestions despite my somewhat dour and taciturn attitude at times. Hopefully the information here can help others in the future.

-Mike

 

[Spearfoot]

All,

I found this post here:

https://serverfault.com/questions/8...y-recognises-sata-drives-but-refusing-to-reco

Seems like this fella had the exact same troubles I was having. I also finally did hear back from Seagate tech support who verified that all the RMA drives I received were all tested just prior to shipping and they blamed the cable or the controller. Supermicro said that while it COULD be an issue with their board or the controller, they tested these model drives and confirmed they work (albeit at 6GB/s instead of the 12Gb/s they are capable of). So based on advice here, and advice from the vendors, and reading that above post very carefully I decided, since I already have close to 1000 bucks invested in this FreeNAS box, I mind as well pony up another 18 bucks for a higher quality cable.

I purchased some Coboc SFF8482-SATA-0.5M cables. I have always had good luck with their displayport and HDMI cables, and they are reputable. In fact on Newegg, someone commented they had this exact board and a SAS drive they needed connecting and this worked (after also complaining several cables were junk).

So this morning, I hooked it up, and sure enough, there in the BIOS under the LSI Controller physical disk map are my Seagate SAS drives confirmed and function. FreeNAS works again since I re-flashed back to the IT BIOS, and its ready to start building my ZFS array.

So...lesson of my SAS story is...buy high quality cables from a reputable manufacturer. I tried two different bargain bin cables from Amazon, and while they worked good with SATA, they did not work with SAS. Something is wired differently someplace somehow. I may just buy an extra cable to dissect and see EXACTLY how one of these cables is wired different than the other.

Also, these particular Seagate drives to NOT spin up with power attached. Seagate tech support also confirmed this is true - the drive platter will only spin up when the controller is ready to access the platter for data transfer. So if anyone else is troubleshooting, you should be able to see the drives over the data cables.

Thanks to everyone for their patience and suggestions despite my somewhat dour and taciturn attitude at times. Hopefully the information here can help others in the future.

-Mike

Good news! "All's well that ends well."

 

[Jailer]

Hopefully the information here can help others in the future.

Thank you for providing an update and a solution to the problem.