[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference pcbuoa::server_works

Title:Server_works
Moderator:PCBUOA::IS_SYSTEM
Created:Fri Jun 30 1995
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:291
Total number of notes:1063

211.0. "RSM only works with BIOS disabled!" by CHEFS::AYLESBURY_L () Thu Feb 13 1997 15:42

    I am working on a trial project to remotely manage an NT web farm. The
    initial project is for 12 Prioris ZX6200 and for them to be completely
    managed (ie; lights out environment and in the middle of London, so no
    easy access)
    
    For the bottom-up management I have been attempting to use the RSM card
    installed in an EISA slot (5) with the latest? version of RSM software,
    as advertised in this conference (V1.1B). I am using a direct
    null-modem cable (home made) which works and I can connect to the card
    fine. I wouldn't have, mind you, if I hadn't read all the notes about
    it!!
    
    The problem I keep running into is whenever I load the BIOS for the RSM
    card the server refuses to boot anything. The only way I can get WNT or
    QuickLaunch to load is to connect to the RSM card and reset the
    on-board firmware. This drops the connection of course but it does
    continue to boot.
    
    The only way I can get anything to boot reliably is to disable the RSM
    BIOS. This means I can't do remote diskette and hold functions, not
    that I've ever seen these working anyway.
    
    All attempts to move the BIOS to other addresses result in the same
    thing and eventually I couldn't connect to the card as it complained at
    having the wrong BIOS version loaded (3.12) and refused to upload the
    new one (3.14). It gets to 37% complete then issues a write failure
    window and drops the connection.
    
    Does this product actually work, or am I being really thick.
    
    Help!!
    
    Les
    
    By the way, the other management tools such as ManageWorks/ServerWorks
    and Remotely Possible work perfectly, first time.
    
T.RTitleUserPersonal
Name
DateLines
211.1Sounds like an RSM BIOS address conflictPCBUOA::BRACKENThu Feb 13 1997 17:2827
    I have RSM installed on a ZX6 and have not encountered this problem.
    You may have an ISA card that is occupying a large portion of BIOS
    space.
    Can you check that? A good Extended BIOS address to use is D8000.
    Also there is no need to have an IRQ set for RSM until we give you a
    driver for the card.
    If you can get the 3.14 version loaded on another machine, a tip to
    check the Extended BIOS addresses that do not work is:
       . from the reset window of the RSM application keep change the
    EBIOS address.
       . then perform a warm boot and wait to see if the boot completes.
       . if it fails attempt the next address until one is found that
    works.
      . if you find one make it permanent with the SCU.
      . if it still fails let me know in this conference withe a
    description of wether it came up with the RSM start banner on the
    console before failing and if it failed at the time of attempting to
    read the boot device. Can you boot from floppy?
    
    The reason it stops in the programming of RSM EBIOS at a certain % is
    that the protocol has pushed that much to the board before it realized
    it can not get at the extended BIOS space that it is trying to program
    into.
    
    Please keep us informed,
    
    Rich   
211.2SUTRA::16.192.160.105::BatsSpeeding, speeding, I'm always speedingMon Feb 17 1997 14:427
	
	We had several off these problems as well when were using
	a Null modem. Replaced it by a real modem, and everything worked.
	(This was during Cebit preparation, and another roadshow some time
	 later)

	Pjotrr
211.3All BIOS tried- same resultCHEFS::AYLESBURY_LWed Feb 19 1997 08:2724
    Rich,
    
    I have tried your suggestion of changing EBIOS addresses but I get the
    same result whatever I use. The only way anything will boot (i.e. get 
    passed the hardware check phase) is to set "no bios address" from the
    RSM window and reset the firmware. This is only good when I can
    actually connect to the card. Most times when this happens it is
    impossible to connect as I keep getting a message claiming the EBIOS
    rev is too low and needs updating. This update always fails at 37%
    complete. I've tried 2 separate cards, same result.
    
    I used the SCU to check the available resourses and there is 64K spare
    at D8000 so that's where I configured it first. It may be I missed
    something in the set-up/install phase. All I've seen is the RSM hardware
    installation guide and that's quite sketchy.
    
    We need to use null modem connections as we plan to connect via reverse
    LAT terminal servers over a dedicated WAN link. The eventual plan is to
    manage hundreds of servers in the same location.
    
    Any other suggestions?
    
    Les
                                       
211.4More to checkPCBUOA::BRACKENWed Feb 19 1997 12:1832
    
>    I have tried your suggestion of changing EBIOS addresses but I get the
>    same result whatever I use. The only way anything will boot (i.e. get 
>    passed the hardware check phase) is to set "no bios address" from the
>    RSM window and reset the firmware. This is only good when I can
>    actually connect to the card. Most times when this happens it is
>    impossible to connect as I keep getting a message claiming the EBIOS
>    rev is too low and needs updating. This update always fails at 37%
>    complete. I've tried 2 separate cards, same result.

     Can you upgrade the EBIOS on some other Server(non ZX6)?
    
>    I used the SCU to check the available resourses and there is 64K spare
>    at D8000 so that's where I configured it first. It may be I missed
>    something in the set-up/install phase. All I've seen is the RSM hardware
>    installation guide and that's quite sketchy.
     
     You can still have a conflict with an ISA card. SCUs can not check on
     ISA card conflicts.
     Remove any ISA cards.
     Where exactly does the Booting stop?
     What is the last message on the console?
     Does it get to RSM EBIOS nessage?
    
>    We need to use null modem connections as we plan to connect via reverse
>    LAT terminal servers over a dedicated WAN link. The eventual plan is to
>    manage hundreds of servers in the same location.
    
     Are you sure the full set of signals are present for H/W Flow Control?
     If you were able to upgrade the RSM ObBoard Firmware via the same cable
     then that is probably not the problem.

211.5In answer?CHEFS::AYLESBURY_LWed Feb 19 1997 15:3221
    I only have an HX5133 to play with although the final installation will
    be into ZX6200MP's
    
    Ther are no other cards in the ISA/EISA slots. All I have is the system
    card with on-board Adaptec SCSI and a PCI Mylex RAID card.
    
    The boot stops at:-
    
    	Quicklaunch--> stops at Loading MS Windows, after showing the
    system details box.
    	WNT----------> stops at NTdetect checking hardware... it normally
    takes about 5 seconds then switches to OS Loader. When it hangs it
    never gets to OS Loader screen.
    
    Yes, the RSM BIOS loaded message appears.
    
    The cable was designed and made after reading all the notes about it! I
    don't have DSR connected though. Does this matter?
    
    Thanks
    Les
211.6DSR should be tied with DTR & CDPCBUOA::BRACKENWed Feb 19 1997 17:3210
    With RSM EBIOS 3.10-11 there was a problem of someone overwriting RSM's
    RAM Memory when CD bootable BIOS came out. This was resolved in RSM1.1
    with the EBIOS 3.14 version moving the stack pointer down further 
    in memory. This should not prevent the upgrade of EBIOS however.
    
    Please add the DSR pin 6 connecting it up with 8(CD) and 20(DTR).
    Try the download again. You never answered the question if the Firmware
    upgrade to 1.1 happened using this same cable?
    
    Rich
211.7Firmware uploads OKCHEFS::AYLESBURY_LThu Feb 20 1997 15:306
    Rich,
    
    I will modify the cable as you suggest and let you know if things
    improve. The firmware uploads fine using the current cable wiring.
    
    Les