Flashing the LSI2308 on SuperMicro X10SL7-F to IT mode. Which version to use NOW?

[danb35]

You don't strictly need to disconnect the drives, but the paranoid among us often do.

...while the more foolhardy among us do the update on a live FreeNAS system with the pool mounted.

 

[Ericloewe]

P.S. There is already alerting ability built into FreeNas for Hard Drive Temps (not sure if you were aware of this)

Apparently it's unreliable. It's on my list of things to check next month...

 

[jgreco]

Apparently it's unreliable. It's on my list of things to check next month...

Also it is only going to work at whatever the SMART polling frequency is.

 

[Mirfster]

Apparently it's unreliable. It's on my list of things to check next month...

Hmm.. Now I got to go get all paranoid... :P

Also it is only going to work at whatever the SMART polling frequency is.

Yeah, guess I should have mentioned that.

 

[Kei]

Seems to me that you may have had more than one instance with overheating if you had to go that far.

Nope, just over cautious :) I run this script also to get a reminder in the morning with the current status, so I know when it's time to adjust the fan controller (in the summer for example). I always try to stay between 30-35 celsius :)

 

[Ericloewe]

Also it is only going to work at whatever the SMART polling frequency is.

Same goes for any alternative, since you can't get the temperature without getting all of -A's output. A separate script doesn't bring anything to the table (assuming FreeNAS' does work...), since you can set the frequency to whatever you want.

 

[jgreco]

Same goes for any alternative, since you can't get the temperature without getting all of -A's output. A separate script doesn't bring anything to the table (assuming FreeNAS' does work...), since you can set the frequency to whatever you want.

Yes, but a lot of UNIX guys have odd ideas about SMART and polling. It should be checked frequently, short tests run often, etc. If you look at many RAID controllers they're polling at 120 or 300 seconds IIRC.

I'd love to see a much smarter physical disk management layer, one that would be aware of scrub/resilver status, and then be launching SMART short tests every four hours, and long tests periodically, so that none of them overlap. It isn't really clear to me why there's such resistance to being aggressive about front-line fault detection, but whatever.

 

[Bidule0hm]

Apparently it's unreliable. It's on my list of things to check next month...

I agree, they never worked on my server.

 

[zoon01]

Supermicro has released now the 20.00.07.00-IT firmware.
See: ftp://ftp.supermicro.com/driver/SAS/LSI/2308/Firmware/IT/

 

[Kei]

Supermicro has released now the 20.00.07.00-IT firmware.
See: ftp://ftp.supermicro.com/driver/SAS/LSI/2308/Firmware/IT/

Nice. Did you upgrade it? I'm cureently running FN9.10.2 with PH20.00.04.00-IT. Do I need to upgrade or are both PH20.00.04.00-IT and PH20.00.07.00-IT compatible with FN9.10.2 ?

 

[zoon01]

Yes, i do have upgraded, but i do not use FreeNAS, i use NAS4Free 11 series.
You should have no problems with it it on your system.

basically it's the same firmware serie so no problems only some fixes included.

Those are the fixes between 20.00.04.00 and new 20.00.07.00:

Description Of Change:

1.There was a check in sending the target missing event that was failing because the location of the initiator/target changed. This check was nuanced to only care about that if the device is only a target.

2.There is a check missing when sending Target Add events up to the host. It does not check if the device had a target remove previously sent. Since the device was no longer missing, the standard checks to prevent this situation were passed.

3.Modified firmware source code such that when an expander is missing from the topology, the Device Present bit in the Flags attribute of SAS Device Page 0 will not be set.

4.When the false reply parity error occurs, clear out the interrupt and return. If it was a real parity error, then the error would either be seen upstream or downstream as well, and another parity error would be generated.

 

[Kei]

Thank you zoon. I guess I will wait some time and then gather some feedback from the community to see is someone flashed this version.... I dont want to be the first one to do it :D

 

[Spearfoot]

Thank you zoon. I guess I will wait some time and then gather some feedback from the community to see is someone flashed this version.... I dont want to be the first one to do it :D

I flashed my X10SL7-F's LSI controller to the new version 20.00.07.00 today. So far, so good.

EDIT: But to be honest, you can probably stay at version 20.00.04.00 and be just fine. I've been running my system with the older version forever, with no problems whatsoever. So go ahead, procrastinate!

 

[Jailer]

I flashed my X10SL7-F's LSI controller to the new version 20.00.07.00 today. So far, so good.

EDIT: But to be honest, you can probably stay at version 20.00.04.00 and be just fine. I've been running my system with the older version forever, with no problems whatsoever. So go ahead, procrastinate! :)

I still am on 20.00.04.00

 

[Ericloewe]

I flashed my X10SL7-F's LSI controller to the new version 20.00.07.00 today. So far, so good.

Well, it's not really new, it's something like half a year old. It fixes a few weird edge cases, so it's definitely recommended for new flashes or if there are weird issues. I wouldn't bother updating from .04 if no issues are present, though.

 

[Kei]

I flashed my X10SL7-F's LSI controller to the new version 20.00.07.00 today. So far, so good.

EDIT: But to be honest, you can probably stay at version 20.00.04.00 and be just fine. I've been running my system with the older version forever, with no problems whatsoever. So go ahead, procrastinate! :)

Yeah, maybe I will upgrade nex time I power down the NAS for the next cleaning :)
However, please allow me to ask you something. I remember that I've had a lot of troubles inserting the SAS address (the last 9 digit you're supposed to insert). What happened to me was that I couldn't type anything, and if I just typed enter, the system returned me a weird error. At this point it might be useful to mention that I run the UEFI version, not the DOS one. Anyways, I dont remember what I did, but I sort of remember that I edited the SMC2308T.NSH file, inserting "sas2flash.efi -o -sasaddhi 5003048xxxxxxxxx" (with the correct last 9 digits) instead of just "sas2flash.efi -o -sasaddhi 5003048". Is it possible that you found the same problem?

 

[Spearfoot]

Yeah, maybe I will upgrade nex time I power down the NAS for the next cleaning :)
However, please allow me to ask you something. I remember that I've had a lot of troubles inserting the SAS address (the last 9 digit you're supposed to insert). What happened to me was that I couldn't type anything, and if I just typed enter, the system returned me a weird error. At this point it might be useful to mention that I run the UEFI version, not the DOS one. Anyways, I dont remember what I did, but I sort of remember that I edited the SMC2308T.NSH file, inserting "sas2flash.efi -o -sasaddhi 5003048xxxxxxxxx" (with the correct last 9 digits) instead of just "sas2flash.efi -o -sasaddhi 5003048". Is it possible that you found the same problem?

Yes, I got a PAL error when I first tried to run the DOS version on my X10SL7, so I've used a UEFI shell ever since.

RE: Setting the SAS address -- I always edit the script and specify the full SAS address, but you have to change the command option from -sasaddhi to -sasadd to do this. They're slightly different: the first prompts you for the remaining SAS address digits, the second lets you specify the complete address. So, you want something like this:

Code:

echo -off
sas2flash.efi -o -e 7
sas2flash.efi -f 2308T207.ROM
sas2flash.efi -b mptsas2.rom
sas2flash.efi -b x64sas2.rom
sas2flash.efi -o -sasadd 5003048xxxxxxxxx

 

[Kei]

Very well explained, thank you! One last thing, is it true that you could remove the "sas2flash.efi -b mptsas2.rom" line of code if using with FreeNAS?

 

[Spearfoot]

Very well explained, thank you! One last thing, is it true that you could remove the "sas2flash.efi -b mptsas2.rom" line of code if using with FreeNAS?

You're welcome!

Yes you can omit that command to flash the ROM. You generally don't need it unless you're booting from the card, or using it as a RAID controller instead of in IT mode.

 

[Kei]

Gotcha:)