Re: [Scsirastools-developers] New user (Fujitsu Sun drives)

[Cress, A. R <and...@in...>]

That's weird, the slot status being read in step 2 does not show that
the disk fault bit has been turned on for disk 0.  Perhaps the HSBP only
uses its own leads to set the read status and doesn't look at the disk
fault bit that was in the write status.  Maybe there is another way to
read that.  I'll look into it. =20
It would definitely show a different status if the disk(s) were really
removed or not ready, though. =20

I need to go ahead and release a new 1.5.2 version with the changes made
so far.
Thankfully, the case of more than one disk fault (especially while still
ready) at the same time should be rare, so I think that change can wait
until 1.5.3.

Andy=20

-----Original Message-----
From: Daniel Smolik [mailto:ma...@my...]=20
Sent: Thursday, March 15, 2007 9:17 AM
To: Cress, Andrew R; scs...@li...
Subject: Re: [Scsirastools-developers] New user (Fujitsu Sun drives)

Cress, Andrew R napsal(a):
> Hmmm.  It has to set all of the slot LEDs on a backplane when it sets
> any one of them, but it reads the current status first before changing
> the one indicated.
>=20
> Can you send me the sgsafte.log with -x turned on for that case? =20
> That should show what the last read status is for the slots. =20
>=20

Yes,
I do :

  ./sgsafte -x -d0 -f > /tmp/log


  ./sgsafte -x -d1 -f >> /tmp/log
./sgsafte -x -d0 -f >> /tmp/log







> Andy=20
>=20
> -----Original Message-----
> From: Daniel Smolik [mailto:ma...@my...]=20
> Sent: Wednesday, March 14, 2007 3:57 PM
> Cc: Cress, Andrew R; scs...@li...
> Subject: Re: [Scsirastools-developers] New user (Fujitsu Sun drives)
>=20
> Daniel Smolik napsal(a):
>=20
>>Cress, Andrew R napsal(a):
>>
>>
>>>Dan,
>>>
>>>OK, I have posted to the SVN trunk an updated sgsafte.c with more
>>>sophisticated handling for >=3D2 HSBP proc devices like you have.
>>>Please see if this handles the disk fault LEDs better, and send me
>=20
> the
>=20
>>>debug log.=20
>>
>>
>>Hi,
>>there is debug log now. Status of LEDs I can test tomorrow.
>>Displaying informations about disks and processor are now  excelent
>=20
> :-)
> I today test LED on enclosure. Situation is better. I can switch  to=20
> orange each led this is good. But if I turn on  LED1 and after this a=20
> try turn on LED 2, LED 1 turn off and LED2 turn on. If I try turn on
LED
>=20
> 3 LED 2 turn off and LED 3 turn on. I cannot get all LEDs on.
>=20
>=20
> 			Regards
> 					Dan


--=20
Mydatex s r.o.
http://www.mydatex.cz
email: sm...@my...
mob: 604200362
tel: 226210085