Conference bulova::decw_jan-89_to_nov-90

RE: .0

 I wanted to make things more clear. The VAXstation is M38
SPX. The problem occurs always after invoking DECdecision
(even if I did not quit from the DECdecision application).

 I have logged through the console (OPA0:) after I did
not get the DIGITAL logo and the login box, I have discoverd 
that the server process is gone. I did not see any thing in 
it's error log or output file.

				*-Doron-*

    When you installed the SPX, did you do an AUTOGEN?
    

The SPX is quiet new. It was installed few weeks ago.
We R going to have here the biggest computer exhibition
ever made in Israel called Digital90. This W.S. (among 
other) was installed from scratch, (on another SPX and 
then backuped image to this particular machine). After
the backup was made an AUTOGEN was issued. Should we
check for some particular parameters?

We R quiet short in time since next week the show begins
it's final preparations.


	Thanks in advance,
			*-Doron-*

    If you log in after the error occurs, can you locate a
    DECISION$*_BUGCHK.DMP file anywhere -- probably in SYS$LOGIN of the user
    who was running DECdecision.  Check the creation date/time to see if it
    corresponds roughly with the time at which the error occurred.
    
    If so, post it in the DECdecision notes file, VIA::DECDECISION.

There are some DECISION$*_BUGCHK.DMP files that match
the date. BUT... I just got reports that some other SPXs
are "suffering" from a similiar problem with out running
DECdecision...

I can post the dumps into the DECdecision notes file.

Are there special sysgen parameters that I have to 
take in concideration?

		*-Doron-*


				

Well, I am back again. DECdecision people could not find 
any thing wrong with the dumps. They claim that there is
a DECwindows problem. Any way, the problem STILL exists
and we are quiet desparate. If no one can help (DECw.'s
engineering) may be I should submit SPR/QAR?

I did set accounting and as I remember, DECW$SERVER_0
termination was access vilolation. (I can not recheck
because accounting is disabled for disk space reasons).

			*-Doron-*

QAR this problem if you want DECwindows engineering attention.  Remember,
the majority of the readers of this conference are NOT DECwindows engineers!

Please include a copy of DECW$SERVER_0_ERROR.LOG in your QAR.  Thank you.

>
>QAR this problem if you want DECwindows engineering attention.  Remember,
>the majority of the readers of this conference are NOT DECwindows engineers!
>
The problem in QAR is that some times the response is far from be quick. I need
a solution/workaround by next week!!!!

			*-Doron-*

If the server died by an accvio, I find it very surprising that there is no
interesting traceback in the log file.  Are you sure you looked at the right
version?  There should be a stack traceback at least.

Burns

How much stuff do you have in MODPARAMS.DAT?  Did you try cleaning out
MODPARAMS.DAT, running AUTOGEN SAVPARAMS REBOOT (after the system has been up
for >24 hours and after you ran into the problem while it was up)?

-Mike

RE:-.1

 Modparams has just new values for LRPCOUNT and NPAGEDYN.
By this evening (local time in Israel) I'll have the 
work station on my table, (I hope), and it will make 
life easier to check other things.
				*-Doron-*

I'll try to put it all togther:
SYS$SYSROOT:[SYSMGR]DECW$SERVER_0_ERROR.LOG;2

 5-APR-1990 13:22:05.1 Hello, this is the X server
Dixmain address=13074
Now attach all known txport images
%DECW-I-ATTACHED, transport DECNET attached to its network
%DECW-W-ATT_FAIL, failed to attach transport LAT
-SYSTEM-F-PRIVINSTALL, shareable images must be installed to run privileged imag
e
in SetFontPath
Connection 99700 is accepted by Txport
out SetFontPath
ScanProc color support loaded
scn$InitOutput address=12acd0
Connection Prefix: len == 60
 5-APR-1990 13:22:40.6 Now I call scheduler/dispatcher
 5-APR-1990 13:22:42.0 Connection 99738 is accepted by Txport
 5-APR-1990 13:22:44.7 Connection 99700 is closed by Txport
 5-APR-1990 13:23:06.6 Connection 99738 is closed by Txport
 5-APR-1990 13:23:10.3 Connection 99700 is accepted by Txport
 5-APR-1990 13:23:13.1 Connection 99770 is accepted by Txport
 5-APR-1990 13:23:19.3 Connection 99738 is accepted by Txport
 5-APR-1990 13:24:33.7 Connection 9adf8 is accepted by Txport
 5-APR-1990 13:25:26.5 Connection 9ae30 is accepted by Txport
 5-APR-1990 13:25:58.2 Connection 9ae68 is accepted by Txport
 5-APR-1990 13:26:00.7 Connection 9aea0 is accepted by Txport
 5-APR-1990 13:26:02.1 Connection 9aed8 is accepted by Txport
 5-APR-1990 13:26:02.7 Connection 9af10 is accepted by Txport
 5-APR-1990 13:27:12.8 MEMMGR/XREALLOC - zero amount is requested
 5-APR-1990 13:27:13.1 MEMMGR/XREALLOC - zero amount is requested
 5-APR-1990 13:27:18.9 Connection 9ae30 is closed by Txport
 5-APR-1990 13:27:19.5 Connection 9ae68 is closed by Txport
 5-APR-1990 13:27:20.2 Connection 9aea0 is closed by Txport
 5-APR-1990 13:27:25.5 MEMMGR/XREALLOC - zero amount is requested
 5-APR-1990 13:27:25.8 Connection 9af10 is closed by Txport
 5-APR-1990 13:27:27.8 Connection 9aed8 is closed by Txport
 5-APR-1990 13:28:23.2 Connection 99738 is closed by Txport
 5-APR-1990 13:28:23.6 Client 3 resets the server
  5-APR-1990 13:28:25.8 FreeAllUnusedFonts: gc_ref_count = 1 for SYS$COMMON:[SYS
FONT.DECW.75DPI]FIXED.DECW$FONT;1
 5-APR-1990 13:28:26.4 Destroying Loadable Microcode Context
 5-APR-1990 13:28:26.7 ...Deallocating Ucode ROM
 5-APR-1990 13:28:27.1 ...Deallocating Ucode FIXED
 5-APR-1990 13:28:29.2 ScanProc color support loaded
scn$InitOutput address=12acd0

SYS$SYSROOT:[SYSMGR]DECW$SERVER_0_OUTPUT.LOG;1


scn_FreeVisual
scn_closeScreen

Accounting:

DETACHED Process Termination
----------------------------

Username:          SYSTEM            UIC:               [SYSTEM]
Account:           SYSTEM            Finish time:        5-APR-1990 13:28:32.51
Process ID:        0000011E          Start time:         5-APR-1990 13:22:03.03
Owner ID:                            Elapsed time:                0 00:06:29.48
Terminal name:                       Processor time:              0 00:00:28.02
Remote node addr:                    Priority:          6
Remote node name:                    Privilege <31-00>: 17110805
Remote ID:                           Privilege <63-32>: 00000000
Queue entry:                         Final status code: 1000000C
Queue name:
Job name:
Final status text: %SYSTEM-F-ACCVIO, access violation, reason mask=!XB, virtual

Page faults:             4266        Direct IO:               1365
Page fault reads:         138        Buffered IO:             1116
Peak working set:        2808        Volumes mounted:            0
Peak page file:          7107        Images executed:            1

MODparams:

SCSSYSTEMID   = 64624  !Don't change it to zero value
SCSNODE       = "bot11a"
VAXCLUSTER    = 0
WINDOW_SYSTEM = 1      !1 for DECwindows    2 for VWS
BUGCHECKFATAL = 0      !Bug in VMS Ver 5.2 autogen take 1 insted of
MIN_SRPCOUNT=1024
MIN_LRPCOUNT=32
MIN_pagedyn=650000
PAGEFILE=0
SWAPFILE=0
DUMPFILE=0

			*-Doron-*

From watching the error log I have noticed that :
%DECW-W-ATT_FAIL, failed to attach transport LAT
-SYSTEM-F-PRIVINSTALL, shareable images must be installed to run privileged imag

I figured out that DECW$TRANSPORT_LAT is the image. After
installing it (by force) things seemed to be OK!

I have tried to disable LAT at boot time and things were OK
again!!! But, we NEED the LAT!!! I can NOT disable the
@SYS$MANAGER:LTLOAD from the systartup_v5.com. I know 
that I suffer from lack of knowledge (trying to improve)
but how do I get the DECW$TRANSPORT_LAT to be installed
(not by specifying in the INSTALL utility)?

				*-Doron-*

Failure to attach to LAT transport should not cause what you are seeing.

    In fact, the LAT transport is not supported for use by the server at
    this time, so you should simply remove it from the list of transports
    to use (see DECW$PRIVATE_SERVER_SETUP).
    
    What I thought was interesting was that the ACCVIO did not appear in
    your server's log file.  USER mode ACCVIOs should be written into the
    log file, along with traceback information.
    
    James

Thanks for all of those who replied. I do not want to discuss
about the LAT transport now. The urgent thing is why do I 
have this problem?
The funny thing is, that now I ave put things back to 
normal, (LAT going up & the DECW$TRANSPORT_LAT is NOT
installed) and I DO NOT have the problem. The
DECW$SERVER_ERROR.LOG does let me know about the 
privilleged image that should be installed, but the server
does not die on quiting. Any one can come and say that 
I can be happy now, but since this work station id going
out for the biggest computer show ever made in Israel 
(and it happend to be that Digital is the one that is 
putting it up), I can not let ANY thing wrong goes 
there.

There is NO DECW$PRIVATE_SERVER_SETUP.COM in SYS$MANAGER.
So where can the problem 'attack' from?

I am quite short in time, but any help would be appreciated.
Since I did not installed the work station, and never installed
VMS 5.3-1, is there any importance to the order of installing
VMS5.3-1 & DECwindows?

			thnx i nadvance,
					*-Doron-*

Hi,

 After more investigations I have played further more with DECdecision 
(drawing charts, using the builder etc...). I quit NORMALy from the application
 so No dumps were made. I have quit from the session and AGAIN, the logo & the 
 login box did not come back. What bothers me is that the DECW$SERVER_0_ERROR.LOG
 does not day any thing but in the accounting, an ACCESS VILOLATION is the reason
 for the proecess termination. There is no reason mask and PC supplied there.
 I have noticed that each time I quit from the session, 2 error logs are created.
 The lower version is empty, and higher version (probably the server tries to 
 be recreated again(?)) contains the log that was enclosed before.

		Any ideas?
				*-Doron-*

    This was asked previously, but have you looked at the workstations's
    resources and quotas?  I've seen so many problems reported that were
    due to insufficient resources and quotas, that I've developed the
    personal philosophy to give the workstation user just about everything
    that is available.  After all, a workstation is essentially a personal
    VAX, so why limit resource usage like I would on a larger shared VAX.
    I suspect that not everyone would agree with the generous values below,
    but I've attached after the form feed some of the resource and quota
    values that I use on our workstations here.  Maybe you would want to
    try them and see if they help.  Don't forget to AUTOGEN after making
    the MODPARAMS changes.

              Dave
    
    UAF quotas

        MaxJobs:         0  Fillm:       100  Bytlm:       100000
        MaxAcctJobs:     0  ShrFillm:      0  Pbytlm:           0
        MaxDetach:       0  BIOlm:       100  JTquota:       2048
        Prclm:          12  DIOlm:       400  WSdef:          500
        Prio:            4  ASTlm:       800  WSquo:         3100
        QuePrio:         0  TQElm:        80  WSextent:      7200
        CPU:        (none)  Enqlm:       600  Pgflquo:      40000

    MODPARAMS.DAT

	! DECwindows related parameters
	WINDOW_SYSTEM=1
	! User specified VMS parameters
	MIN_GBLPAGFIL=15000
	MIN_DEFMBXBUFQUO=2048
	MIN_DEFMBXMXMSG=2048
	MIN_GBLPAGES=45000
	MIN_GBLSECTIONS=400
	MAXPROCESSCNT=50
	MIN_PQL_DWSDEFAULT=300
	MIN_PQL_MWSDEFAULT=200
	MIN_PQL_DWSQUOTA=500
	MIN_PQL_MWSQUOTA=300
	MIN_PQL_DWSEXTENT=800
	MIN_PQL_MWSEXTENT=500
	MIN_PQL_DBYTLM=32768
	MIN_PQL_MBYTLM=32768
	MIN_PQL_DDIOLM=100
	MIN_PQL_MDIOLM=100
	MIN_PQL_DBIOLM=100
	MIN_PQL_MBIOLM=100
	MIN_PQL_DENQLM=400
	MIN_PQL_MENQLM=200
	MIN_PQL_DASTLM=400
	MIN_PQL_MASTLM=200
	MIN_PQL_DFILLM=100
	MIN_PQL_MFILLM=60
	TTY_ALTALARM=1498
	TTY_ALTYPAHD=1648
	TTY_TYPAHDSZ=1648

Thnx Dave,

 Since now it's Passover in Israel ( I am working from home via the MACintosh),
I'll try your parameters and report back in this notes file. I realy hope it
helps...
 What I do miss, is guidlines that will instruct tuning for DECwindows (some
thing like chapter 5 in the 'old VWS' installation guide. Can some one
point/write down such instructions?

				*-Doron-*

Dave, I am sorry to inform U that although U have tried to help,
with the enclosed MODPARAMS.DAT & UAF parameters, it di not help!!!

I have autogened and reboot but the problem still occurs!!!

Any other (*URGENT*) sugestion?

		*-Doron-*