Hi,

The SMS address space is one of the Top 10 I/O consumers on most z/OS systems 
here.  It is small compared to DB2 and IMS but large compared to most other 
persistent tasks.

Has anyone ever had a conversation with IBM or raised a requirement to reduce 
the physical I/O to the COMMDS when the SMSPLEX is identical to the Sysplex?  
Maybe this is a dumb question and there is some obvious tuning knob I have been 
missing.   I noticed we do have interval set to 10 seconds instead of the 
default 15 and it has been that way apparently since we implemented SMS in the 
1990's.  When I was looking at this it just struck me as really odd that almost 
every other product we have that that does checkpoint processing to communicate 
system status can optionally exploit the coupling facility or XCF but not 
DFSMS.  I keep an eye on Top I/O consumers looking for any changes and usually 
skipped over SMS but I had a "Why is the sky blue?" moment today. 

JES2, Thruput Manager, CA-MIM all exploit XCF or Coupling Facility structures 
to exchange status information and reduce data set "checkpoint" I/O.

D SMS,OPTIONS                                            
IGD002I 19:54:14 DISPLAY SMS 756                         
ACDS     = SYS2.SMS.ACDS                                 
COMMDS   = SYS2.SMS.COMMDS                               
INTERVAL = 10           DINTERVAL = 150                  
SMF_TIME = YES          CACHETIME = 3600                 
CF_TIME = 3600          PDSE_RESTARTABLE_AS = YES        
PDSE_BMFTIME = 3600     PDSE1_BMFTIME = 3600             
PDSE_LRUTIME = 60       PDSE1_LRUTIME = 60               
PDSE_LRUCYCLES = 15    PDSE1_LRUCYCLES = 15              
LOCAL_DEADLOCK = 15     GLOBAL_DEADLOCK = 4              
REVERIFY = NO           DSNTYPE = PDS                    
ACSDEFAULTS = NO        PDSESHARING = EXTENDED           
OVRD_EXPDT = NO         SYSTEMS = 8                      
PDSE_HSP_SIZE = 0MB   PDSE1_HSP_SIZE   = 0MB             
USE_RESOWNER = YES      RLS_MAX_POOL_SIZE = 100MB        
RLSINIT = YES           RLSTMOUT = 0                     
CICSVR_INIT = NO        COMPRESS = GENERIC               
CICSVR_DSNAME_PREFIX = DWW.                              
CICSVR_RCDS_PREFIX = DWW                                 
CICSVR_GRPNAME_SUFFIX = PROD                             
CICSVR_ZZVALUE_PARM =                                    
CICSVR_UNDOLOG_CONTROL =                                 
CICSVR_UNDOLOG_PREFIX = DWW                              
CICSVR_BACKOUT_CONTROL =                                 
CICSVR_GENERAL_CONTROL =                                 
Rls_MaxCfFeatureLevel = A                                
RlsAboveTheBarMaxPoolSize = 0                            
RlsFixedPoolSize = 0                                     
PDSE_MONITOR = (YES,0,0)  PDSE1_MONITOR = (YES,0,0)      
PDSE_DIRECTORY_STORAGE = 2000M                           
PDSE1_DIRECTORY_STORAGE = 2000M                          
PDSE_BUFFER_BEYOND_CLOSE = YES                           
PDSE1_BUFFER_BEYOND_CLOSE = YES                          
GDS_RECLAIM = YES       DSSTIMEOUT = 0                   
BLOCKTOKENSIZE = NOREQUIRE       FAST_VOLSEL = OFF       
IGD002I 19:54:14 DISPLAY SMS                             
TRACE    = OFF    SIZE = 128K     TYPE = ERROR           
  JOBNAME = *    ASID = *                                
  TRACING EVENTS:                                        
    MODULE = ON   SMSSJF = ON   SMSSSI = ON   ACSINT = ON
    OPCMD  = ON   CONFC  = ON   CDSC   = ON   CONFS  = ON
    MSG    = ON   ERR    = ON   CONFR  = ON   CONFA  = ON
    ACSPRO = ON   IDAX   = ON   DISP   = ON   CATG   = ON
    VOLREF = ON   SCHEDP = ON   SCHEDS = ON   VTOCL  = ON
    VTOCD  = ON   VTOCR  = ON   VTOCC  = ON   VTOCA  = ON
    RCD    = ON   DCF    = ON   DPN    = ON   TVR    = ON
    DSTACK = ON   UAFF   = ON   DEBUG  = ON              
VOLSELMSG = (OFF,0)    TYPE = ERROR    JOBNAME = *       
    ASID = *  STEPNAME = *                               
    DSNAME = *                                           


Anyone have any thoughts on the list or privately if you don't want to comment 
on IBM-MAIN is fine too?


        Best Regards, 

                Sam Knutson, GEICO 
                System z Performance and Availability Management 
                mailto:[EMAIL PROTECTED] 
                (office)  301.986.3574     
         
"Think big, act bold, start simple, grow fast..." 





====================
This email/fax message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution of this
email/fax is prohibited. If you are not the intended recipient, please
destroy all paper and electronic copies of the original message.

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html

Reply via email to