Re: [OT] Clock of cpu1 (and more) is shown incorrectly at boot in multiprocessor in -current

[Erik Fair <fair%netbsd.org@localhost>]

On Jun 4, 2014, at 00:44 , BERTRAND Joël <joel.bertrand%systella.fr@localhost> 
wrote:

>       I see you have two HyperSPARC. Is your SS20 stable ? I have two 
> RT626@200 Mhz in mine and it regulary panics with -current. I can see a lot 
> of ESP errors also (I have _two_ internal discs).

You're worried about heat and stressing the power supply, I'd guess. The other 
thing is that while the internal disks are on "wide" (16-bit) SCA connectors, 
they're not hooked up "wide" to the SCSI controller, so they max out at 10MB/s 
channel (of course, the disk may or may not be able to deliver in excess of 
that, depending upon how old they are) - if you want the 20MB/s performance one 
can get from wide SCSI, one has to hook up the disks in something like Sun 611 
cabinets to a wide Sbus SCSI controller. That will also remove them from the 
cabinet (reducing heat load) and remove them from the power supply load.

        Erik <fair%netbsd.org@localhost>