[Info-vax] Pinning down a SYSTEM-F-EXQUOTA

[Jan-Erik Soderholm]

Hi.

I have a batch job running a Python script that exits with
SYSTEM-F-EXQUOTA when the Python code reads in a "to large"
JPG file. I can "solve" it by reducing the size of the JPG
file, but I'd also liked to understand what is happening.

A 211 Kbyte file runs OK then it crashed on a 311 Kbyte file.

The error seems to happen when the Python scripts write a
Base64 encoded vesion of the JPG over the net in an XML
transaction using SUDS (a web-service tool for Python).

The error :
%TYPE-F-WRITEERR, error writing SYS$OUTPUT:.;
-RMS-F-SYS, QIO system service request failed
-SYSTEM-F-EXQUOTA, process quota exceeded
$

I have mainly played with BYTLM, WSEXT and PGFLQUO, both
in UAF and as PQL_Mxxx params. No solution yet.

$ tcpip sh ver

   HP TCP/IP Services for OpenVMS Alpha Version V5.6 - ECO 5
   on an AlphaStation XP900 466 MHz running OpenVMS V8.3

512 Mb memory.

I have tried with SHO PROC/CONT "Q" in another session and that
exits with the last display :

                              Process BATCH_82                     14:46:33



     State               COM                 Working set               2921

     PID                 000010D9            Page faults               7608

     UIC                 [SYSTEM]            Event flags           C3000001
                                                                   C8000000

     # open files remaining               501/512                  ( 97%)
     Direct I/O count/limit              1024/1024                 (100%)
     Buffered I/O count/limit            1024/1024                 (100%)
     BUFIO byte count/limit           3998272/3998272              (100%)
     ASTs remaining                       298/300                  ( 99%)
     Timer entries remaining              399/400                  ( 99%)
     PGFL quota count/limit            996695/1000000              ( 99%)
     ENQ quota count/limit               2000/2000                 (100%)

     $1$LDA2:[python254.][vms.bin]python.exe;1


$

The rellevant parts of ACCOUNTING says :

Queue entry:       81                Final status code: 00000001
Queue name:        TRADSCHED_BATCH
Job name:          ADD_ITEM_2
Final status text: %SYSTEM-S-NORMAL, normal successful completion
Page faults:             7000        Direct IO:               1471
Page fault reads:         301        Buffered IO:             8092
Peak working set:       91536        Volumes mounted:            0
Peak page file:        317920        Images executed:            8

The user UAF record says :

Maxjobs:         0  Fillm:       512  Bytlm:      4000000
Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0
Maxdetach:       0  BIOlm:      1024  JTquota:      60000
Prclm:           0  DIOlm:      1024  WSdef:        20000
Prio:            4  ASTlm:       300  WSquo:       200000
Queprio:         0  TQElm:       400  WSextent:   1200000
CPU:        (none)  Enqlm:      2000  Pgflquo:   16000000

Is there some particular setting that limits large transfers ?
Such as in TCPIP maybe ?

Jan-Erik.