| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 6019.1 | | DECWET::SCHREIBER | DECeNT | Tue May 20 1997 15:12 | 2 |
| The Session Manager either did not start, or started and terminated
abnormally.
|
| 6019.2 | Yes, I agree. | STAR::HREAD | | Wed May 21 1997 10:05 | 16 |
|
I believe I understood the session manager couldn't load. What I really need to
know is WHY the session manager couldn't load so I can fix it.
The failure status code that was issued with the session5_init_fail error was
0XF0000522. If someone has access to a copy of INIT.C, perhaps they would be
willing to look up the cause of this crash and post the information here so I
can fix the system and get it to boot.
I would really appreciate that.
Thanks,
Harold
|
| 6019.3 | | DECWET::SCHREIBER | DECeNT | Wed May 21 1997 12:39 | 11 |
| I told you exactly what's happening in the code. The code does a
create process, and then waits for 5 seconds for the Session Manager to
start (this was the process created). If the process isn't there, you
get session5_fail.
Why didn't the session manager load? Damn good question. Debugging
setup is very complex, and generally can't be done without help from
the setup developers at Microsoft.
Have you tried another CD? Is the disk IDE or SCSI? What is the
Noname Alpha clone, anyhow? Have you called the supplier for support?
|
| 6019.4 | | STAR::HREAD | | Wed May 21 1997 13:54 | 25 |
| Ah, now I understand better. The CD I'm trying to load is a beta V4.0 server CD.
I've also tried to load a V3.5 CD. It didn't work, either. I don't have any
other CDs to try. The CD did load on a 486 processor machine, though.
The CD and hard disks are both SCSI. Noname is the code name for the AXP PCI
motherboard product offered by Digital. There is a notes file for AXPPCI. I
looked there and then here for information related to the problem I'm having.
Support for this board is supposed to be included on the CD. It does appear to
have the appropriate files for the AXPPCI product on the CD. It appears that the
startup gets pretty far along past where the processor specific stuff is set up,
so I figure the problem is a generic one, not a processor specific one. That's
why I posted a help request here rather than in the AXPPCI notes file.
Does the failure status code that was issued with the session5 failure message
help narrow down the choices for a possible cause for the session manager
failure?
Where does the session manager code get loaded from, the CD or the hard disk?
What other state is the software in at this point in the startup process?
Does NT use a swap file or something like that?
|
| 6019.5 | Is 5 seconds enough (what speed is your 21066?) | BBPBV1::WALLACE | john wallace @ bbp. +44 860 675093 | Sat May 24 1997 05:59 | 23 |
| Sounds like we're talking AXPpci33 here (a 21066 typically at 166 or
233 MHz on a baby-AT board). A number of different clock speeds and
cache sizes can be selected (by jumpers). Is it conceivable that your
config needs more than 5 seconds to get the session manager up and
running ?
Given the right config of Noname this should work - I'm pretty sure I
know at least one person who has NT 4 (non beta) working on AXPpci33.
If he's not using 4 it's retail 3.51. Are you sure this board is
supported by 3.5 ? Either way, it won't be real quick...
Is your firmware up to date ? (Not sure why it would cause this, but..)
Yes there's a page file, eventually (but is there one at this stage?).
At this stage I *think* most of the files are being loaded from the
temporary directory on the system disk, rather than from the CD.
I'd check speed-related jumper settings, and then try to get hold of a
final 3.51 or 4 CD if at all possible.
regards
john
|