Conference csc32::consolemanager

Nand,

  I am going to be really honest with you, I dont really know
  what your problem is, except that for others who have seen
  this, increasing quotas has removed the problem.

Cheers,
Phil

    Please post the results of a 
    
    SHO TERMINAL TTA2:
    
    while PCM is running.
    
    
    Regs,
      Dan

Terminal settings when VCS is stopped and Console Manager started:

Terminal: _TTA2:      Device_Type: Unknown       Owner: Console Ctrl 01
                                              Username: SYSTEM
 
Terminal Characteristics:
   Interactive        No Echo            Type_ahead         No Escape
   Hostsync           TTsync             Lowercase          Tab
   No Wrap            Hardcopy           No Remote          Eightbit
   No Broadcast       No Readsync        Form               Fulldup
   No Modem           No Local_echo      No Autobaud        No Hangup
   No Brdcstmbx       No DMA             Altypeahd          No Set_speed
   No Commsync        No Line Editing    Insert editing     No Fallback
   No Dialup          No Secure server   No Disconnect      Pasthru
   No Syspassword     No SIXEL Graphics  No Soft Characters No Printer Port
   Numeric Keypad     No ANSI_CRT        No Regis           No Block_mode
   No Advanced_video  No Edit_mode       No DEC_CRT         No DEC_CRT2
   No DEC_CRT3        No DEC_CRT4        DEC_CRT5           No Ansi_Color
   VMS Style Input

    Looks good! I set up basically the same environment on my system except
    I used TTA3 on both the PCM host and the 4000VLC workstation serviced
    node and it works fine! We'll have to track down where this message
    is signalled and where that code is coming from. It doesn't match any
    message in any message file I can find.
    
    Dan

    
    
    Phil,Dan,
    
    Any news on this error code ?
    Which quota(s) do you mean (UAF ?,PQL ? ,...) ?
    
    Rgds,
    
    Geert

	Hi Phil, Dan,

	I took over de problem from Nand.
	It think it's time to tell to whole story.

	Customer's configuration and history :
	
	VCS V1.4,OpenVMS VAX 5.5-2
	VCS system VAXstation 3100 monitors 9 systems.
	One of these systems (a PC) has a direct connection 
	to the VCS system(tta2).
	Works fine. Happy customer.

	Customer wanted OpenVMS VAX 6.1. So PCM is needed.
	We did the upgrade to VMS V6.1, installed PCM V1.5 with ECO kit and
	migrated VCS to PCM.
	If he starts PCM, he has problems with the direct connection(tta2) :
Multi line window events window:
 "Console link for system PC_XMS lost - Unrecognized system error code: 5816
Connection lost to managed system "
	After some time he gets :
"Connection re-established to managed system"
	
	And this continues.

	Under VCS he doesn't has this problem.

	SHOW TERMINAL under VCS
	-----------------------
Terminal: _TTA2:      Device_Type: Unknown       Owner: VCS$IODL
                                              Username: SYSTEM

Terminal Characteristics:
   Interactive        Echo               Type_ahead         No Escape
   Hostsync           TTsync             Lowercase          No Tab
   No Wrap            Scope              No Remote          No Eightbit
   No Broadcast       No Readsync        No Form            Fulldup
   No Modem           No Local_echo      No Autobaud        No Hangup
   No Brdcstmbx       No DMA             Altypeahd          Set_speed
   No Commsync        Line Editing       Insert editing     No Fallback
   No Dialup          No Secure server   No Disconnect      Pasthru
   No Syspassword     No SIXEL Graphics  No Soft Characters No Printer Port
   Numeric Keypad     No ANSI_CRT        No Regis           No Block_mode
   No Advanced_video  No Edit_mode       No DEC_CRT         No DEC_CRT2
   No DEC_CRT3        No DEC_CRT4        DEC_CRT5           Ansi_Color
   VMS Style Input

	SOW TERMINAL under PCM
	----------------------
Terminal: _TTA2:      Device_Type: Unknown       Owner: Console Ctrl 01
                                              Username: SYSTEM

Terminal Characteristics:
   Interactive        No Echo            Type_ahead         No Escape
   Hostsync           TTsync             Lowercase          Tab
   No Wrap            Hardcopy           No Remote          Eightbit
   No Broadcast       No Readsync        Form               Fulldup
   No Modem           No Local_echo      No Autobaud        No Hangup
   No Brdcstmbx       No DMA             Altypeahd          No Set_speed
   No Commsync        No Line Editing    Insert editing     No Fallback
   No Dialup          No Secure server   No Disconnect      Pasthru
   No Syspassword     No SIXEL Graphics  No Soft Characters No Printer Port
   Numeric Keypad     No ANSI_CRT        No Regis           No Block_mode
   No Advanced_video  No Edit_mode       No DEC_CRT         No DEC_CRT2
   No DEC_CRT3        No DEC_CRT4        DEC_CRT5           No Ansi_Color
   VMS Style Input



	What can we do the help the customer ? He is not happy
	with this situation(6 weeks now) and for the moment he likes VCS
	more than PCM. We must convince the customer about PCM.
	Therefore any help is welcome.

	Thanks in advance.

	Geert

    My problem is that I cannot reproduce it. I see nothing wrong with the
    terminal settings. Off hand, the other thing I can think of to check 
    are the settings of tty_altypahd and tty_altalarm. What are they?
    
    Regs,
      Dan

  Another one to check is that the line isnt asserting modem signals, we
  should tell you that its hung up, but never having tried it I dont know.

Cheers,
Phil

	Hi Phil,Dan,

SYSGEN>  SHO TTY_ALTYPAHD
Parameter Name            Current    Default     Min.     Max.     Unit  Dynamic
--------------            -------    -------    -------  -------   ----  -------
TTY_ALTYPAHD                 1000        200         0     32767 Bytes

SYSGEN>  SH TTY_ALTALARM
Parameter Name            Current    Default     Min.     Max.     Unit  Dynamic
--------------            -------    -------    -------  -------   ----  -------
TTY_ALTALARM                  320         64         0        -1 Bytes


                   X25 switch
                  --------
       PC         |      |
      -----       |      |
      |   |       |      |                                  VAX
      |	  |-------|      |                                ---------
      |	  |	  |      |                                |       |
      -----	  |      |RS-232 output               TTA2|       |
		  |      |--------------------------------|       |
		  --------                                |       |
							  |       |
							  ---------        


	B.T.W. If we look at sh term tta2 (VCS versus PCM), we see that 
    	VCS tta2 has "No Eightbit" and PCM has "Eightbit". Could this be
    	the reason ?(I don't think,but ...)
    
    
    	Rgds,
    
    	Geert

    Modem is turned off plus the tt ports on modern vaxstations are DZ
    based and don't even have full modem control much less there isn't
    enough wires in the small MMJ connectors to provide the necessary
    signals. I cannot see eightbit .vs. no eightbit causing this problem as
    this  effect how certain esc sequences are processed. I did not know
    there was a "switch" involved here. WHat I would like to know now is if
    the RS232 port on the switch is setup for XON/XOFF flow control. yes I
    know VCS works but again it handled reads from ports much differently
    than PCM does. It is worth checking.
    
    Regards,
       Dan

	Hi Dan,

	This X25 switch RS-232 port has XON,ASCII,7 bit,even parity char's.
	The customer does not see the Error code 5816 anymore.
	BUT via PCM he doesn't see messages via CONSOLE EXTRACT.
	He does see them with VCS review(if he shuts PCM and start 
	VCS of course).
	When he starts CONSOLE MONITOR he gets strange characters in the
	beginning and than nothing ...

	Thanks and rgds,

	Geert

    >7 bit even parity....
    
    
    I think we definitely have a problem here as PCM doesn't have a setting
    for parity in the config databases so your getting whatever the port
    is set too. Of course you will see wierd characters when you have a parity
    mismatch. I'm betting you've got 8 bits and no-parity on the port.
    
    Regs,
      Dan

    
    
    Dan,
    
    I asked the customer to change(test) the port setting on that
    X25 switch(Begonia Telindus) in the same way as tta2(PCM).
    I will give feedback.
    Stay tuned.
    
    Rgds,
    
    Geert

	Hi Dan,

	Eventually I had feedback from the customer(he didn't want to cooperate
	anymore...).
	I asked him to make a direct connection between a VAX and his PCM system
	via tta2: . Guess what. Yes that works fine.
	The customer removed the VAX and made again the connection from the PC
	to the X25 switch and from the X25 switch(RS-232 port) to tta2: of his
	PCM system. As to him both X25 switch RS-232 port and tta2 have
	8 bits and no-parity set. 
	For him it's not working as expected -->output he gets:

 ��30����9����:53:����No�e����H��PD�SKݬ��i�e�����el90�_3�6�5_�05�ݬ�St�t��Ko.
 ��30����9����:5�:����No�e����H��PD�SKݬ��i�e�����el90�_3�6�5_�05�ݬ�St�t����.
 ��30����9����:55:0���No�e����H��PD�SKݬ��i�e�����el90�_3�6�5_�05�ݬ�St�t��Ko.
 ��30����9����:56:00��No�e����P�����AMݬ��i�e��3��el0��_��6_�65ݬ�St�t��Ko.
 ��30����9����:56:0���No�e����H��PD�SKݬ��i�e�����el90�_3�6�5_�05�ݬ�St�t����.
 ��30����9����:5�:0���No�e�03�B��_3ݬ���P�3��St�t���묠���e��S�.
 ��30����9����:5�:�3��M�jor��No�e�0��B��_�ݬ��i�e��0�D�S_�N_ݬ�St�t��Ko.
 ��30����9����:5�:����M�jor��No�e�03�B��_3ݬ��i�e��0�D�S_BU_�U�_N�Kݬ�St�t����.
 ��30����9����:5�:05��No�e����P�����AMݬ��i�e��3��el0��_��6_�65ݬ�St�t����.
 ��30����9����:59:�0��No�e�03�B��_3ݬ���P�3��St�t���묠���e��S�.
 
��30����9����:06:�3��M�jor��No�e�05�ANV_�ݬ��i�e����D�_MA����A���ݬ�St�t���묠�
��
��e�0000�.
 ��30����9����:06:5���No�e�03�B��_3ݬ���P�3��St�t���묠���e��S�.
 ��30����9����:��:5���M�jor��No�e�0��B��_�ݬ��i�e��0�D�S_�N_ݬ�St�t����.
 ��30����9����:��:59��M�jor��No�e�03�B��_3ݬ��i�e��0�D�S_BU_�U�_N�Kݬ�St�t��Ko.
 ��30����9����:�6:�0��No�e�03�B��_3ݬ���P�3��St�t���묠���e��S�.


	And with VCS he gets :

PC_XMS 23/11/94,08:17:25, Major, Node=03[BXL_3], Line=34[RAF_DESK_TEST_PORT], St
atus=Ok.
PC_XMS 23/11/94,08:19:10, Node=41[ASLK_CGER], Line=12[bel260_486_402], Status=Ko
.
PC_XMS 23/11/94,08:19:50, Node=45[IPPA], Line=12[bel331_486IN_446], Status=Ko.
PC_XMS 23/11/94,08:20:19, Node=33[BBL_MARNIX], Line=13[bel282_486M_203], Status=
Ko.
PC_XMS 23/11/94,08:20:40, Node=41[ASLK_CGER], Line=12[bel260_486_402], Status=Ok
.
PC_XMS 23/11/94,08:20:43, Node=33[BBL_MARNIX], Line=13[bel282_486M_203], Status=
Ok.
PC_XMS 23/11/94,08:21:10, Node=45[IPPA], Line=12[bel331_486IN_446], Status=Ok.
PC_XMS 23/11/94,08:21:52, Node=10[DEWAAY_10], Line=11[bel008_486_640], Status=Ko
.
PC_XMS 23/11/94,08:23:07, Node=10[DEWAAY_10], Line=11[bel008_486_640], Status=Ok
.
PC_XMS 23/11/94,08:23:29, Node=11[PETERCAM], Line=15[bel024_486_1007], Status=Ok
.

	Thanks and rgds,

	Geert
                  

    
    
    	Hi Dan,Phil,
    	
    	I have spoken with a communication specialist--> I think this is
    	a better picture of the problem:
    
                              |-------------|
                              |             |
      PCM(tta2) -------------pad1----------pad2-----------PC
                              |             |
                              |-------------|
                                 X 25 switch
    
    	The customer changed the tta2 and pad1 characteristics(ie. 8 bit...)
        But of course he must change also these for pad2, which he can't do.
    	Reason to much applications have to be changed.
    	With VCS there was no problem because everything was 7 bit.
    
    	The customer is not happy at all because he had VCS(7 bit -->works fine)
    	and PCM(8 bit hardcoded ? -->problem).
    
    	Can engineering change this in the next release. Meaning that
    	both 7 bit/8 bit is possible.
    
    
    	Thanks and Rgds,
    
    	Geert 
    
    	

Geert,
      I have been looking at this with Phil today, and now your explaination
clears things up for me.

We will do something to help the customer for the MUP/ECO kit we are preparing
at the moment.

I do however feel that the problem does not lie with PCM.

You have two PADS at each end of the X.25 switch and you tell the PADS
what data format to expect on the asych connection, but from what you said
the PAD is ignoring the fact that you set one PAD up as 7 bit and the other as
8 bit. From my thinking it should be a function of the PAD to put out the data
in the format you told it i.e. in your case converting 7 to 8 bit  and vice
versa.

What we are looking to do as a work around for the customer is to have a logical
which when defined will cause PCM to set ALL it's LAT and Direct devices to 7
bit mode. 

Cheers Simon..

	Hi Simon,

>      I have been looking at this with Phil today, and now your explaination
>clears things up for me.
 	Great
>We will do something to help the customer for the MUP/ECO kit we are preparing
>at the moment.
 	Great.Great. When will it be available ? Customer wants us to 
	open a CLD for this.
>I do however feel that the problem does not lie with PCM.
 	I understand what you mean. But the customer got a point,
	because with VCS no problem.
>You have two PADS at each end of the X.25 switch and you tell the PADS
>what data format to expect on the asych connection, but from what you said
>the PAD is ignoring the fact that you set one PAD up as 7 bit and the other as
>8 bit. From my thinking it should be a function of the PAD to put out the data
>in the format you told it i.e. in your case converting 7 to 8 bit  and vice
>versa.
 	I don't know. This is not my speciality. I will check with
	my communication collegue...
>What we are looking to do as a work around for the customer is to have a logical
>which when defined will cause PCM to set ALL it's LAT and Direct devices to 7
>bit mode. 
 	Great.Great.Great.
	What did you specify 8 bit for PCM and not 7 bit like for VCS ?

	Rgds,

	Geert

    >        Great.Great.Great.
    >        What did you specify 8 bit for PCM and not 7 bit like for VCS ?
    
    I need to clear up a few things in my mind here as I worked on serial
    commuciations for years and I have a few issues with whats being said
    here.
    
    First of all, when you 7 or 8 bit what are you really trying to say?
    Are you talking about character size and parity or what? If you 
    mean the setting of the Eightbit or Noeightbit that one sees from SHOW
    TERMINAL then we're talking about a totally different animal. That
    terminal setting controls whether how 8 bit ESC sequence will be
    handled. It has nothing to do with  the parity of the comm line. 
    I haven't looked at the IODL code in a while but as I remember VCS
    didn't mess with the parity setting either, however it would indeed
    trap a parity error and ignore it. I'll check the code, but again
    I think it did a SENSECHAR and just kept the parity setting of the line
    whenever it did the SETCHAR and SETMODE.
    
    I'm not jumping on anyone, I just want to make sure we undertsand the
    real problem here. I'll take a look at the IODL code and determine
    exactly what it did.
    
    
    Regs,
      Dan

Dan,

>>    I'm not jumping on anyone, I just want to make sure we undertsand the
>>    real problem here. I'll take a look at the IODL code and determine
>>    exactly what it did.

  We appreciate any and all the help we can get here, Simon and I spent the
whole of Friday trying to understand and fix this problem with differing
bit/parity etc. So, your "years" of experience would certainly be of benefit.

Cheers,
Phil

	Hi Dan,Phil,Simon,

>    First of all, when you 7 or 8 bit what are you really trying to say?
>    Are you talking about character size and parity or what? If you 
>    mean the setting of the Eightbit or Noeightbit that one sees from SHOW
>    TERMINAL then we're talking about a totally different animal. That
>    terminal setting controls whether how 8 bit ESC sequence will be
>    handled. It has nothing to do with  the parity of the comm line. 

	Indeed, I meant the terminal characteristics are different
	between VCS(8 bit) and PCM(7 bit). 
	On the PADS, the custmer has 7 bit and parity even.

    
>    I'm not jumping on anyone, I just want to make sure we undertsand the
>    real problem here. I'll take a look at the IODL code and determine
>    exactly what it did.
	I'm not a communication specialist. I'm only trying to explain
	the customer's problem.
    
    
   	Thanks and Rgds,
        
	Geert

Geert,

  Can you ask your customer to try the new CONSOLE$DAEMON.EXE_VAX image
  I have placed in OPG::CM$KIT:

  In order for this to work correctly, they will have to define a
  couple of logicals, CONSOLE$PRIVATE_SETUP.COM would be a good
  place for these...

  $ Define/TABLE=CONSOLE$LOGICAL_NAMES CONSOLE$NOEIGHTBIT TRUE
  $ Define/TABLE=CONSOLE$LOGICAL_NAMES CONSOLE$PARITY xxxx

    Note: xxxx should be either ODD or EVEN depending on what your
          customer requires.

  Please let us know if this works ASAP (we need to know by wednesday) so
  that we can put it in the MUP kit we are in the process of building.

  I will be doing something a little better for V2.0, I will probably
  allow these values to be set on a system by system basis in the PCM
  database.

Cheers,
Phil

    
    Hi Phil,
    
    Good news. The customer did the test and it works fine.
    
    Thank you very much.
    
    Rgds,
    
    Geert

Geert,

  Great news, please have your customer install the MUP when it is released,
  this will ensure that this fix plus any others we include between now and
  release time are being used.

Cheers,
Phil

    Well thats what I get for taking a day off and not reading notes before
    researching this...;-}
    
    For the record, VCS didn't muck with parity and it set the necessary
    bits to disable parity errors and to pass characters to the reader 
    even if a parity error occurred.
    
    So VCS works becuase it didn't mess with anything. If you set parity
    then it kept it when the IODL setup the terminal lines.
    
    Regs,
      Dan

    Dan,
        what you take a day off - Never!! :-))
    
    It does seem rather interesting that VCS did not muck with the data
    type (7 or 8 bits) and nor did we, but they appeared to use 7 bits and
    we used eight. I don't know whether anything changed between V5 and V6
    in the default settings for the terminal chars.
    
    Cheers Simon...

    Simon,
      This was still bugging me a little bit so I did some more digging. I
    think what we had was a combination of parity and the setting of
    eightbit with regards to how characters are interpreted on the read
    from the line. When I mentioned 8 it an esc sequences in an earlier
    note I should have qualified it better. Typically I find people getting
    bit by this on reads of esc sequences where the CSI character is used.
    CSI is an 8 bit code and *cannot* be received if we have a parity
    setting since the parity bit takes up the space for it. In our case
    we were reading in eighbit mode and accepting the characters with the 
    eigth bit set and logging them. In essence, we are operating with the
    8-bit ASCII character set. When a decterm or VT terminal displays
    them it displays a character in the 8-bit ascii character set hence the
    unintelligable garbage displayed by the PCM user interfaces. With the
    line set for 7-bit characters, the 8th bit is *discarded*. VCS worked
    with Geert's customer setup because it setup the line to ignore all parity
    errors and it left the setting of eightbit alone (which is noeightbit
    by default). Also note that on writes the setting of eightbit is
    irrelevant.
    
    Some other points:
    
    Systems that transmit 8-bit ascii codes (to perform screen formatting
    for example) should always be setup to use 8-bits/no parity, otherwise
    PCM will get only the first 7 bits of the character. Also, parity
    settings are ignored on LAT lines so if you need a particular parity
    you have to setup the physical terminal server ports themselves for
    that parity.
    
    
    Regs,
      Dan

    A customer reported also this error 5816. He got it when he
    disconnected a cable.
    It is clear that this is not the normal way of life, but the error
    message isn't clear either.
    
    Paul