| Title: | VAX and Alpha VMS |
| Notice: | This is a new VMSnotes, please read note 2.1 |
| Moderator: | VAXAXP::BERNARDO |
| Created: | Wed Jan 22 1997 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 703 |
| Total number of notes: | 3722 |
Hi
I have already posted this question in Alphanotes (#69),but with no good answer
surely because i don't post clearly my question, so I reformulate it here.
For the AS1000 (MIKASA) and the AS1000A (NORITAKE) they exist the
following machinechecks:
225 mchk$c_ps1fail Power supply 1 failed in redundant configuration
227 mchk$c_ps2fail Power supply 2 failed in redundant configuration
Question:
When will be a new sys$laodable_images:sys$cpu_routinesxxx for AS1000 and
AS1000A will be modified to include these two new machinecheck codes in
order to VMS to make an entry in errlog/console to inform you
"eh! power supply # is dead"
Actually with up to VMS V6.2-1H3 (i have not tested with V7.1) you have a
problem when one power deads.
As you are in redundant mode VMS continues to run (this is OK)
but you are not informed that one power is bad (no entry in errlog, no
message at the console) up you power off power on the system (or init it
under the >>>).
Only at that time you are informed that one power is bad, and the console
shutdowns the other power and you can't continue.
see below an message console example
>>>self test...
** PWR S 2 failed **
** Interrupt mask disabled **
** Shutting power down **
only at that time you know one power is bad, but it's too late, customer
is obliged to wait that us, DIGITAL replace the power to continue !
If the information was there before the power up test/init a replacement
power should already been ordered with no loss of time and no customer
anger.
Thank you
Henri
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 405.1 | Contact Mikasa Engineering Team | XDELTA::HOFFMAN | Steve, OpenVMS Engineering | Tue Apr 01 1997 14:37 | 38 |
Please see the MIKASA notes I referenced over in VAXAXP::AlphaNotes 69.1, where I first saw this question. I expect you will have to code this yourself -- I thought that much was clear from the notes you likely already found. -- If you want this sort of platform-specific diagnostic support added to OpenVMS, please contact the revelent platform engineering team with the request, and be prepared to make a case for making this change. (OpenVMS does not normally handle these sorts of events without the input and the assistance of the processor engineering team...) -- Based on a check of the V7.1 OpenVMS listings, the MCK$K_PS1FAIL and MCK$K_PS2FAIL signals, as well as MCK$K_TEMPWARN and various other signals, can be accepted by OpenVMS from the PALcode. And OpenVMS does have code to call the make_errorlog routine present in module [CPU1B02]ERROR_ROUTINES_1B02.C, in routine exe$system_corrected_error, when the mck$l_pal_errcode value received from the PALcode indicates this is required. If it's not clear from the above, this support is entirely based on information received from the PALcode, and the Mikasa team is responsible for the PALcode... (If you are running OpenVMS Alpha V7.1 and are not seeing these messages on power supply failure, then you'll want to check with the Mikasa engineering team to start to figure out why...) -- (And if you're asking these questions, I'd acquire a copy of the source listings for OpenVMS -- ordering information is present in note 8.* or 9.* in VAXAXP::VMSNOTES. The internal price of the source listing CD-ROM is cheap, and is well worth the subscription charge.) | |||||
| 405.2 | Thanks a lot | PRSSOS::FONDI | Wed Apr 02 1997 02:32 | 4 | |
Thanks a lot, Steve for your enlighted and detailed and cleared reply.
Now, i know how "bad power" event must be reported to VMS...
Henri
| |||||