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

Conference bulova::decw_jan-89_to_nov-90

Title:DECWINDOWS 26-JAN-89 to 29-NOV-90
Notice:See 1639.0 for VMS V5.3 kit; 2043.0 for 5.4 IFT kit
Moderator:STAR::VATNE
Created:Mon Oct 30 1989
Last Modified:Mon Dec 31 1990
Last Successful Update:Fri Jun 06 1997
Number of topics:3726
Total number of notes:19516

2936.0. "Compute bound DECW$SERVER process" by CSSE32::CARR (Trust everybody, but cut the cards.) Thu Jun 14 1990 17:58

	I'm having problems with the DECW$SERVER process going compute
	bound. I'm fairly sure that it's a process quota, sysgen
	parameter, or DECW setup issue, since the problem only occurs
	during periods of heavy activity (typically at login time).

	The VMS version is T5.4-4GE. I have several systems (each with
	identical quotas/parameters/setups) exhibiting the symptom, so it's
	not h/w related (all systems are VS3100 with SPX/VRT19).

	Below I've appended the output of a SHO PROC/CONT/ID=mumble. While
	this runs NOTHING on the display ever changes except the compute
	time and the PC (which toggles from 15CB08 to 15CB00).

	I've also appended the code segment around the PC. This shows a
	simple CMPL/BRx loop, which is not going to be exited without some
	sort of AST to release it(?).

	I've also appended the result of a SDA> SHOW PROCESS, which shows
	adequate quotas remaining....

	I urgently need to get this fixed in order to take the systems to
	DECworld, so any help would be much appreciated.

	Nick
	




                             Process DECW$SERVER_0              16:33:01



    State               COM                 Working set              904

    Cur/base priority   6/6                 Virtual pages           6133

    Current PC          0015CB08            CPU time     000:00:30:22.86

    Current PSL         03C00009            Direct I/O              1418

    Current user SP     7FF91884            Buffered I/O            1492

    PID                 00000090            Page faults             7057

    UIC                 [SYSTEM]            Event flags         60000001
                                                                80000000


    N450II$DKA200:[SYS0.SYSCOMMON.][SYSEXE]DECW$SERVER_MAIN.EXE;1



SDA> exa/instr 0015CB08-20;30
0015CAE8:  REMQHI  @-5271(R0),@-2153(SP)
0015CAEF:  CHMK    #10
0015CAF1:  MOVAL   04(AP),R0
0015CAF5:  CMPL    @00(R0),#00004000
0015CAFD:  BGEQU   0015CB0A
0015CAFF:  NOP
0015CB00:  CMPL    @00(R0),#00004000
0015CB08:  BLSSU   0015CB00
0015CB0A:  TSTL    08(AP)
0015CB0D:  BEQL    0015CB14
0015CB0F:  MOVL    08(AP),@0C(AP)
0015CB14:  RET
0015CB15:  TSTL    R0
0015CB17:  NOP
0015CB18:  HALT



Process index: 0010   Name: DECW$SERVER_0   Extended PID: 00000090
------------------------------------------------------------------
Process status:  00140011   RES,PSWAPM,PHDRES,LOGIN

PCB address              803E3A70    JIB address              805FAB80
PHD address              80A07E00    Swapfile disk address    00000000
Master internal PID      00010010    Subprocess count                0
Internal PID             00010010    Creator internal PID     00000000
Extended PID             00000090    Creator extended PID     00000000
State                       COM      Termination mailbox          0000
Current priority                6    AST's enabled                KESU
Base priority                   6    AST's active                 NONE
UIC                [00001,000004]    AST's remaining                96
Mutex count                     0    Buffered I/O count/limit       58/60
Waiting EF cluster              0    Direct I/O count/limit         99/100
Starting wait time       19001919    BUFIO byte count/limit      46160/46160
Event flag wait mask     DFFFFFFF    # open files allowed left      83
Local EF cluster 0       60000001    Timer entries allowed left      7
Local EF cluster 1       80000000    Active page table count         0
Global cluster 2 pointer 00000000    Process WS page count         776
Global cluster 3 pointer 00000000    Global WS page count          128
T.RTitleUserPersonal
Name
DateLines
2936.1Need more informationSTAR::VATNEPeter Vatne, VMS DevelopmentThu Jun 14 1990 18:102
Please append the output of an SDA SHOW PROCESS/IMAGES command,
so we can relate the PC to a particular image.
2936.3...as requested in .1CSSE32::CARRTrust everybody, but cut the cards.Fri Jun 15 1990 07:5724
SDA> sho proc/images
Process index: 0010   Name: DECW$SERVER_0   Extended PID: 00000090
------------------------------------------------------------------


                            Process activated images
                            ------------------------

ICB       Start     End       Type            Image Name  Major ID,Minor ID
--------  --------  --------  --------------  -----------------------------
7FFCBB00  00000200  000007FF  MAIN            DECW$SERVER_MAIN  0,0
7FFCBCC0  00004200  000555FF  GLOBAL          DECW$SERVER_DIX  4,6
7FFCBB70  00098800  000AF7FF  GLOBAL     SHR  VAXCRTL  4,3
7FFCBC50  00055600  0007F5FF  GLOBAL     SHR  MTHRTL  129,32780
7FFCBBE0  0007F600  000987FF  GLOBAL     SHR  LIBRTL  1,14
7FFCBD30  00000800  000041FF  GLOBAL PRT SHR  DECW$TRANSPORT_COMMON  12,12
7FFCBE80  00110800  001123FF  MERGED PRT SHR  DECW$TRANSPORT_DECNET  0,0
7FFCBF60  00114400  00114FFF  MERGED          DECW$TRANSPORTMSG  0,0
7FFCBEF0  00115000  00115DFF  MERGED     SHR  VAXCMSG  0,0
7FFCBDA0  00115E00  001179FF  MERGED PRT SHR  DECW$TRANSPORT_LOCAL  0,0
7FFCBE10  0014D200  001639FF  MERGED          DECW$SERVER_DDX_GE  0,0

Total images = 11		Pages allocated = 1624
SDA> exit
2936.4Looks like an SPX problemSTAR::VATNEPeter Vatne, VMS DevelopmentFri Jun 15 1990 12:526
From what I can tell, the PC is smack in the middle of the SPX
dependant code.  There have been bug fixes for serious SPX problems
lately.  I don't know if one of those will solve your problem.

However, since this is urgently needed for DECworld, we are
handling this problem off-line.
2936.5CSSE32::CARRTrust everybody, but cut the cards.Fri Jun 15 1990 14:063
    Thanks, the patches seem to have fixed it....
    
    Nick
2936.6Same problem V5.3-2DIODE::CROWELLJon CrowellFri Jun 15 1990 20:076
    
    I have the same problem with VMS 5.3-2 on my SPX.  Could someone
    point me to the correct patches..  
    
    Thanks,
    Jon
2936.7DECWIN::ROSENBLUMMon Jun 18 1990 10:224
The V5.3-1 patches are the correct ones for all 5.3-*
releases.

Mike
2936.8Could someone give a pointerDIODE::CROWELLJon CrowellTue Jun 19 1990 01:249
    
    Are the patches on the net?  I've never needed to get any patches in
    the past and wouldn't know where to look. 
    
    The server takes off about once every few days now, it takes about
    5 min's log in and kill the process. 
    
    Thanks,
    Jon
2936.9SPX patches please!SALSA::MARSHALLWed Jun 27 1990 15:2816
>               <<< Note 2936.6 by DIODE::CROWELL "Jon Crowell" >>>
>                            -< Same problem V5.3-2 >-

    
>    I have the same problem with VMS 5.3-2 on my SPX.  Could someone
>    point me to the correct patches..  
    
	I have the same problem with VMS 5.3-1 on my 36 SPX stations.
	I desperately seeking these patches for 36 SPX stations which
	are part of an Investment Trading Floor system using Digitals
	Price Watch software.  The systems are going into this state
	frequently and we have had to reboot to recover.  The system
	is going live next Monday July 2nd.  I'm in a time crunch --
	help please!

					Bruce Marshall
2936.10Here's where to get them - CSC!MS3100::SCHELLMark Schell, Sls Support, 367-4040Thu Jun 28 1990 14:1227
>================================================================================
>Note 2936.9             Compute bound DECW$SERVER process                 9 of 9
>SALSA::MARSHALL                                      16 lines  27-JUN-1990 14:28
>--------------------------------------------------------------------------------
>                            -< SPX patches please! >-
>
>>    point me to the correct patches..  
>    
>        I have the same problem with VMS 5.3-1 on my 36 SPX stations.
>        I desperately seeking these patches for 36 SPX stations which
>        are part of an Investment Trading Floor system using Digitals
>        Price Watch software.  The systems are going into this state
>        frequently and we have had to reboot to recover.  The system
>        is going live next Monday July 2nd.  I'm in a time crunch --
>        help please!
>
>                                        Bruce Marshall

	Bruce,
		The ONLY way to get the patches is via the CSC!  That's both
the best way and most likely the only supported way to get patches to released
software.

Call them up and get them to copy them across the net to your system, or have
them FedEx it (charged to your cost center, of course) to the customer!

Mark
2936.11CSCPAT_0183010 thats the one!SALSA::MARSHALLWed Jul 04 1990 13:4319
The patch available from the CSC is CSCPAT_0183010.  It does correct the above
problems indicated in Mr. Carr's original note.   I have observed one additional
problem since I applied the patch.  When a user selects the PAUSE feature, 
then enters their password and clicks OK, the windows return as white siloettes
on a dark background and the mouse is dead.  Shortly after, less than 
one minute, the system will crash.  

I am now looking for a way to disable the pause feature, until V5.4 of VMS and
DECWindows V3.0. are available.  Even with this problem, it is STILL MUCH BETTER
to be running with the patch and not experience the crashes we were haveing.


We have been live at the bank Investment Trading floor for two days with no
additional window related problems!


				Thanks,

				Bruce
2936.12Patch to fix the patch?SMOOT::ROTHIraq needs lawyers... send some NOW!!Wed Oct 17 1990 09:1811
I'm running V5.3-1 and have been experiencing the problem outlined in this
note during system startup.

I was ready to call the CSC and get the above mentioned patch but I don't
like what I read about the pause function crashing the system.

Is there any fix for the problem of the pause crashing the system?

Thanks-

Lee