Re: DFSORT and SAS
That Weekly TREND file has been growing, probably a couple decades. It is quite possible that my former SORT product was due to fail for the same general reasons. > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Scott Barry > Sent: Monday, January 25, 2021 2:45 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > It appears that you are managing the //TREND PDB as a SAS sequential- > format (indicated with RECFM=U, and as you mention with prior TMM- > deployment - also getting zEDC benefit, hopefully) and I expect that SAS > does not detect the observation/row/record count which is why you see the > ICE118I message. This is clearly a SAS-limitation situation, what with the > tape-format SAS data library. > > Regards, > Scott Barry > SBBTech LLC > > > On Mon, 25 Jan 2021 21:04:58 +, Gibney, Dave > wrote: > > >I am sure the ICE118I showed up because of the //SORTDIAG > >My Tape management is CA-7. But, all datasets are on disk. > >JCL follows: > >//TRNDDSNS EXEC MXGSAS,WORK='5000,1000' > >//** > >//* DO MXG WEEKLY DATASET TRENDING PROCESSING > >//** > >//WEEK DD DSN=MXG.DATASETS.WEEK(+0),DISP=OLD > >//TRENDDD DSN=MXG.DATASETS.TREND,DISP=OLD > >//SYSINDD * > > %INCLUDE SOURCLIB(SYS802C); > >//SORTDIAG DD DUMMY > >//DFSPARM DD * > >OPTION FILSZ=E1800 > > > >My Tape management is CA-7. But, all datasets are on disk. > > > >/* COPYRIGHT (C) 1993,1994 MERRILL CONSULTANTS DALLAS TEXAS USA > */ > > /* LAST UPDATED: JUL 1, 1994*/ > > /* */ > > /* THIS MEMBER IS A PART OF THE ALL-YOUR-DATA-SETS-TRACKING- > SYSTEM */ > > /* SEE ADOCDSNS FOR DESCRIPTION OF ITS OTHER COMPONENTS*/ > > /* */ > > /*CHANGE LOG*/ > > /*06/03/94 UPDATED FOR NEW VMXGSUM LOGIC*/ > > > /** > **/ > >OPTIONS NODSNFERR NOVNFERR; > >%VMXGSUM(INVOKEBY=TRNDDSNS, > > INDATA= WEEK.DATASETS > > TREND.TRNDDSNS, > > OUTDATA= TREND.TRNDDSNS, > > DSNLABEL=TRND DSNS: TREND 1415/64 DATASETS, > > SUMBY= DSNAME, > > MIN= FIRST, > > MAX= LAST LASTPRI LASTLV1 LASTLV2 LASTBKP LASTTMS, > > SUM= ACTIVE SPACE1-SPACE6 DAYS1-DAYS6 BACKUPS VOLUMES TAPES, > > OUTCODE= > > %INCLUDE SOURCLIB(IMACZDAT); /* SET ZDATE=TODAY() */ > >OPTIONS DSNFERR VNFERR; > > > >A I understand this, it is just a sort or merge of the newly updated TREND > with the previous week into a new generation. > >I don't know why SAS is not informing DFSORT of the data volume. It should > know the new record count precisely. And should be able to see the raw size > of the previous generation.. > > > >The main dataset is a SAS dataset: > >Data Set Name . . . . : MXG.DATASETS.TREND > >General Data Current Allocation > > Management class . . : MCRSRC01Allocated tracks . : 84,226 > > Storage class . . . : SCRSRC Allocated extents . : 7 > > Volume serial . . . : RSRC10 + > > Device type . . . . : 3390 > > Data class . . . . . : BIG > > Organization . . . : PS Current Utilization > > Record format . . . : FS Used tracks . . . . : 83,488 > > Record length . . . : 27648 Used extents . . . : 7 > > Block size . . . . : 27648 > > 1st extent tracks . : 49218 > > Secondary tracks . : 1000 Dates > > Data set name type : Creation date . . . : 2002/03/25 > >Referenced date . . : 2021/01/25 > >Expiration date . . : ***None*** > > SMS Compressible . : NO > >The other is sequential, but still SAS and many years ago, would have been > tape be for we did the SMS Tape Mount Management stuff: > >Data Set Name . . . . : MXG.DATASETS.WEEK.G0979V00 > > > >General Data Current Allocation > > Management class . . : ATMBKUPSAllocated megabytes : 3,000 > > Storage class . . . : SCDSKTAPAllocated extents . : 2 > > Volume serial . . . : PTAP40 + > > Device type . . . . : 3390 > > Data class . . . . . : TMBKUPLG > > Organization
Re: DFSORT and SAS
t Allocation > Management class . . : ATMBKUPSAllocated megabytes : 3,000 > Storage class . . . : SCDSKTAPAllocated extents . : 2 > Volume serial . . . : PTAP40 + > Device type . . . . : 3390 > Data class . . . . . : TMBKUPLG > Organization . . . : PS Current Utilization > Record format . . . : U Used megabytes . . : 15 > Record length . . . : 0 Used extents . . . : 1 > Block size . . . . : 32760 > 1st extent megabytes: 1500 > Secondary megabytes : 2500 Dates > Data set name type : EXTENDEDCreation date . . . : 2021/01/11 >Referenced date . . : 2021/01/25 >Expiration date . . : ***None*** > SMS Compressible . : YES > > >> -Original Message- >> From: IBM Mainframe Discussion List On >> Behalf Of Sri h Kolusu >> Sent: Monday, January 25, 2021 12:29 PM >> To: IBM-MAIN@LISTSERV.UA.EDU >> Subject: Re: DFSORT and SAS >> >> >> OPTION >> SORTDD=SASS,MSGDDN=SYSOUT,MAINSIZE=MAX,MSGPRT=CRITICAL,NOLIS >> T >> >> Dave, >> >> What type of dataset is the input dataset(ddname SASSIN)? Is it a tape >> dataset? If so what kind of tape management system do you have? RMM ? >> CA-1 ? >> >> PS: Your latest joblog shows that you indeed received ICE118I message (4th >> line in the Sortdiag file) >> >> >> Thanks, >> Kolusu >> DFSORT Development >> IBM Corporation >> >> >> >> >> -- >> For IBM-MAIN subscribe / signoff / archive access instructions, >> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > >-- >For IBM-MAIN subscribe / signoff / archive access instructions, >send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
> But, all datasets are on disk. Dave, If the files are on DISK , DFSORT should be able to get the stats. Can you please send me the COMPLETE joblog to my email offline? I want to check allocation messages for the ddname SASSIN. > My Tape management is CA-7 I guess you meant CA-1(Tape management) where as CA-7 is a job scheduling software. CA-1 implements the ICETPEX exit which would be called by DFSORT to get the attributes of the tape file. Thanks, Kolusu DFSORT Development IBM Corporation -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
Secondary megabytes : 2500 Dates Data set name type : EXTENDEDCreation date . . . : 2021/01/11 Referenced date . . : 2021/01/25 Expiration date . . : ***None*** SMS Compressible . : YES > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Sri h Kolusu > Sent: Monday, January 25, 2021 12:29 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > >> OPTION > SORTDD=SASS,MSGDDN=SYSOUT,MAINSIZE=MAX,MSGPRT=CRITICAL,NOLIS > T > > Dave, > > What type of dataset is the input dataset(ddname SASSIN)? Is it a tape > dataset? If so what kind of tape management system do you have? RMM ? > CA-1 ? > > PS: Your latest joblog shows that you indeed received ICE118I message (4th > line in the Sortdiag file) > > > Thanks, > Kolusu > DFSORT Development > IBM Corporation > > > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
>> OPTION SORTDD=SASS,MSGDDN=SYSOUT,MAINSIZE=MAX,MSGPRT=CRITICAL,NOLIST Dave, What type of dataset is the input dataset(ddname SASSIN)? Is it a tape dataset? If so what kind of tape management system do you have? RMM ? CA-1 ? PS: Your latest joblog shows that you indeed received ICE118I message (4th line in the Sortdiag file) Thanks, Kolusu DFSORT Development IBM Corporation -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
I let this week's run fail and have attached the results with //SORTDIAG DD DUMMY and a SAS PROC OPTIONS; I am not sure I'll open an issue with SAS. My site is working towards shutting down later this year or so, and I have a working solution with the DFSORT option on this job. This job is the only failing job (so far) following my shift to DFSORT. Other things are probably more important to us than pursuing this further. > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Scott Barry > Sent: Tuesday, January 19, 2021 4:36 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > SAS application programs often invoke multiple PROC SORT requests (some > thru SQL-type invocations, or otherwise) and so there really should not be an > OPTION FILESZ=Ennn necessary, given the SAS program-invoked > method (i.e., SAS generates the OPTION and SORT statements.) > > And so you should see evidence in SORT diagnostic output messages where > the SAS/DFSORT request is invoked where you really should see the SORT > statement and any preceding OPTION statement which is where you would > see the file-size (estimated) directive. > > So, if the OP codes a //SORTDIAG DD DUMMY statement and if there is no > OPTION statement followed by the SORT FIELDS=() invocation directive, > then you need to open a SAS support TRACK and be ready to provide the > entire job-output along with the SASLOG, also to include your PROC > OPTIONS; output for their review/diagnosis. Something is not right with the > SAS/DFSORT invocation diagnostics output contributed in this IBM-MAIN > group post. > > Regards, > > Scott Barry > SBBTech LLC > > -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN 1PROC OPTIONS; SAS (R) PROPRIETARY SOFTWARE RELEASE 9.01.01M3P11032004 PORTABLE OPTIONS: _LAST_=_NULL_ LAST SAS DATA SET CREATED APPLETLOC=LOCATION OF JAVA APPLETS ARMAGENT= ARM AGENT TO USE TO COLLECT ARM RECORDS ARMLOC=ARMLOC.LOG IDENTIFY LOCATION WHERE ARM RECORDS ARE TO BE WRITTEN ARMSUBSYS=(ARM_NONE) ENABLE/DISABLE ARMING OF SAS SUBSYSTEMS ASYNCHIO ENABLE ASYNCHRONOUS INPUT/OUTPUT AUTOSAVELOC= IDENTIFIES THE LOCATION WHERE PROGRAM EDITOR CONTENTS ARE AUTO SAVED NOAUTOSIGNON SAS/CONNECT REMOTE SUBMIT WILL NOT AUTOMATICALLY ATTEMPT TO SIGNON NOBATCH DO NOT USE THE BATCH SET OF DEFAULT VALUES FOR SAS SYSTEM OPTIONS BINDING=DEFAULT CONTROLS THE BINDING EDGE FOR DUPLEXED OUTPUT BOTTOMMARGIN=0.000 BOTTOM MARGIN FOR PRINTED OUTPUT BUFNO=2 NUMBER OF BUFFERS FOR EACH SAS DATA SET BUFSIZE=0 SIZE OF BUFFER FOR PAGE OF SAS DATA SET BYERR SET THE ERROR FLAG IF A NULL DATA SET IS INPUT TO THE SORT PROCEDURE BYLINEPRINT THE BY-LINE AT THE BEGINNING OF EACH BY-GROUP BYSORTED REQUIRE SAS DATA SET OBSERVATIONS TO BE SORTED FOR BY PROCESSING NOCAPSDO NOT TRANSLATE SOURCE INPUT TO UPPERCASE CARDIMAGE PROCESS SAS SOURCE AND DATA LINES AS 80-BYTE RECORDS CATCACHE=0NUMBER OF SAS CATALOGS TO KEEP IN CACHE MEMORY CBUFNO=1 NUMBER OF BUFFERS TO USE FOR EACH SAS CATALOG CENTERCENTER SAS PROCEDURE OUTPUT NOCHARCODEDO NOT USE CHARACTER COMBINATIONS AS SUBSTITUTE FOR SPECIAL CHARACTERS NOT ON THE KEYBOARD CLEANUP ATTEMPT RECOVERY FROM OUT-OF-RESOURCES CONDITION NOCMDMAC DO NOT SUPPORT COMMAND-STYLE MACROS CMPLIB= IDENTIFY PREVIOUSLY COMPILED LIBRARIES OF CMP SUBROUTINES TO USE WHEN LINKING CMPOPT=(NOEXTRAMATH NOMISSCHECK NOPRECISE NOGUARDCHECK) ENABLE SAS COMPILER PERFORMANCE OPTIMIZATIONS NOCOLLATE DO NOT COLLATE MULTIPLE COPIES OF PRINTED OUTPUT COLORPRINTING PRINT IN COLOR IF PRINTER SUPPORTS COLOR COMAMID=XMS SPECIFIES THE COMMUNICATION ACCESS METHOD TO BE USED FOR SAS DISTRIBUTED PRODUCTS COMPRESS=YES SPECIFIES WHETHER TO COMPRESS OBSERVATIONS IN OUTPUT SAS DATA SETS CONNECTPERSISTPERSIST CONNECTION TO REMOTE SESSION CONNECTREMOTE=REMOTE SESSION ID USED BY SAS/CONNECT SOFTWARE CONNECTSTATUS SHOW THE CURRENT STATUS OF A SAS/CONNECT UPLOAD OR DOWNLOAD TRANSFER CONNECTWAIT WAIT FOR A SAS/CONNECT RSUBMIT TO FINISH BEFORE ALLOWING FURTHER PROCESSING TO OCCUR COPIES=1 NUMBER OF COPIES TO MAKE WHEN PRINTING CPUCOUNT=6NUMBER OF PROCESSORS AVAILABLE. CPUID PRINT CPU INFORMATION AT BEGINNING OF LOG DATASTMTCHK=COREKEYWORDS CONTROL WHETHER KEYWORDS SUCH AS SET, MERGE, ETC. ARE ACCEPTED AS ONE-LEVEL DATASET NAMES IN THE DATA STATEMENT
Re: DFSORT and SAS
SAS application programs often invoke multiple PROC SORT requests (some thru SQL-type invocations, or otherwise) and so there really should not be an OPTION FILESZ=Ennn necessary, given the SAS program-invoked method (i.e., SAS generates the OPTION and SORT statements.) And so you should see evidence in SORT diagnostic output messages where the SAS/DFSORT request is invoked where you really should see the SORT statement and any preceding OPTION statement which is where you would see the file-size (estimated) directive. So, if the OP codes a //SORTDIAG DD DUMMY statement and if there is no OPTION statement followed by the SORT FIELDS=() invocation directive, then you need to open a SAS support TRACK and be ready to provide the entire job-output along with the SASLOG, also to include your PROC OPTIONS; output for their review/diagnosis. Something is not right with the SAS/DFSORT invocation diagnostics output contributed in this IBM-MAIN group post. Regards, Scott Barry SBBTech LLC On Tue, 19 Jan 2021 21:23:21 +, Gibney, Dave wrote: >Seem like something that SAS should do without being told >My checking of SAS options didn't find the settings other than the ones I >mentioned in original note. > >> -Original Message- >> From: IBM Mainframe Discussion List On >> Behalf Of Martin Packer >> Sent: Tuesday, January 19, 2021 1:09 PM >> To: IBM-MAIN@LISTSERV.UA.EDU >> Subject: Re: DFSORT and SAS >> >> I'd ask over on MXG-L if there are methods for SAS to pass the record >> count and average record length estimate to DFSORT. >> >> Cheers, Martin >> >> Martin Packer >> >> Systems Investigator & Performance Troubleshooter, IBM >> >> +44-7802-245-584 >> >> email: martin_pac...@uk.ibm.com >> >> Twitter / Facebook IDs: MartinPacker >> >> Blog: >> https://urldefense.com/v3/__https://mainframeperformancetopics.com__; >> !!JmPEgBY0HMszNaDT!- >> jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ7JjmRfNQ$ >> >> Mainframe, Performance, Topics Podcast Series (With Marna Walle): >> https://urldefense.com/v3/__https://anchor.fm/marna- >> walle__;!!JmPEgBY0HMszNaDT!- >> jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ5GMlQQ6Q$ >> >> Youtube channel: >> https://urldefense.com/v3/__https://www.youtube.com/channel/UCu_65 >> HaYgksbF6Q8SQ4oOvA__;!!JmPEgBY0HMszNaDT!- >> jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ5uyw0NKQ$ >> >> >> >> From: "Gibney, Dave" >> To: IBM-MAIN@LISTSERV.UA.EDU >> Date: 19/01/2021 20:14 >> Subject:[EXTERNAL] Re: DFSORT and SAS >> Sent by:IBM Mainframe Discussion List >> >> >> >> Thank you. Adding the suggested parm allowed my job to succeed. >> Ultimately, there where 15,890,801 records. >> I looked again and there was not an ICE118I issued in the original job. >> I really don't know how, or what controls SAS (v9) passes to the HOST SORT >> >> Thanks again for the assistance. >> >> > -Original Message- >> > From: IBM Mainframe Discussion List On >> > Behalf Of Sri h Kolusu >> > Sent: Tuesday, January 19, 2021 11:37 AM >> > To: IBM-MAIN@LISTSERV.UA.EDU >> > Subject: Re: DFSORT and SAS >> > >> > > ICE752I 0 FSZ=0 RE IGN=0 C AVG=188 0 WSP=0 E DYN=2313 16352 >> > >> > Dave, >> > >> > When DFSORT is invoked via program, it needs to know how many records >> it >> > is sorting so that it can optimally allocate resources needed to sort >> that >> > data. Since the program is feeding the records DFSORT does not have a >> clue >> > and the resources allocated are not enough to complete the sort and >> hence >> > the message ICE046A 0 SORT CAPACITY EXCEEDED - RECORD COUNT >> 6956799 >> > >> > Generally, DFSORT can automatically determine the input file size. >> However, >> > in a few cases, such as when an E15 supplies all of the input records or >> > when information about a tape data set is not available from a tape >> > management system, DFSORT cannot determine an accurate file size. >> > >> > Your joblog does not show one of the important message. You should >> have >> > gotten an ICE118I message >> > >> > ICE118I UNKNOWN FILE SIZE - RECOMMEND SPECIFYING FILSZ=EN TO >> > REDUCE RISK OF >> > FAILURE OR DEGRADED PERFORMANCE >> > >> > INVALID URI REMOVED >> > >> nter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.icem100/mb00083.htm__;!!JmPEg
Re: DFSORT and SAS
Seem like something that SAS should do without being told My checking of SAS options didn't find the settings other than the ones I mentioned in original note. > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Martin Packer > Sent: Tuesday, January 19, 2021 1:09 PM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > I'd ask over on MXG-L if there are methods for SAS to pass the record > count and average record length estimate to DFSORT. > > Cheers, Martin > > Martin Packer > > Systems Investigator & Performance Troubleshooter, IBM > > +44-7802-245-584 > > email: martin_pac...@uk.ibm.com > > Twitter / Facebook IDs: MartinPacker > > Blog: > https://urldefense.com/v3/__https://mainframeperformancetopics.com__; > !!JmPEgBY0HMszNaDT!- > jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ7JjmRfNQ$ > > Mainframe, Performance, Topics Podcast Series (With Marna Walle): > https://urldefense.com/v3/__https://anchor.fm/marna- > walle__;!!JmPEgBY0HMszNaDT!- > jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ5GMlQQ6Q$ > > Youtube channel: > https://urldefense.com/v3/__https://www.youtube.com/channel/UCu_65 > HaYgksbF6Q8SQ4oOvA__;!!JmPEgBY0HMszNaDT!- > jCPCp1oJgGvSG9Dtcsgk0O8m66pT5PMrB2L1_5At3l0h-ggJZr4QQ5uyw0NKQ$ > > > > From: "Gibney, Dave" > To: IBM-MAIN@LISTSERV.UA.EDU > Date: 19/01/2021 20:14 > Subject:[EXTERNAL] Re: DFSORT and SAS > Sent by:IBM Mainframe Discussion List > > > > Thank you. Adding the suggested parm allowed my job to succeed. > Ultimately, there where 15,890,801 records. > I looked again and there was not an ICE118I issued in the original job. > I really don't know how, or what controls SAS (v9) passes to the HOST SORT > > Thanks again for the assistance. > > > -Original Message----- > > From: IBM Mainframe Discussion List On > > Behalf Of Sri h Kolusu > > Sent: Tuesday, January 19, 2021 11:37 AM > > To: IBM-MAIN@LISTSERV.UA.EDU > > Subject: Re: DFSORT and SAS > > > > > ICE752I 0 FSZ=0 RE IGN=0 C AVG=188 0 WSP=0 E DYN=2313 16352 > > > > Dave, > > > > When DFSORT is invoked via program, it needs to know how many records > it > > is sorting so that it can optimally allocate resources needed to sort > that > > data. Since the program is feeding the records DFSORT does not have a > clue > > and the resources allocated are not enough to complete the sort and > hence > > the message ICE046A 0 SORT CAPACITY EXCEEDED - RECORD COUNT > 6956799 > > > > Generally, DFSORT can automatically determine the input file size. > However, > > in a few cases, such as when an E15 supplies all of the input records or > > when information about a tape data set is not available from a tape > > management system, DFSORT cannot determine an accurate file size. > > > > Your joblog does not show one of the important message. You should > have > > gotten an ICE118I message > > > > ICE118I UNKNOWN FILE SIZE - RECOMMEND SPECIFYING FILSZ=EN TO > > REDUCE RISK OF > > FAILURE OR DEGRADED PERFORMANCE > > > > INVALID URI REMOVED > > > nter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.icem100/mb00083.htm__;!!JmPEg > > > BY0HMszNaDT!9_LGpASIz6lgFy3pDduAe2DC4jQhXQBtSHYnTd10a0weZdx0ie > > 4mdHNVwnuilA$ > > > > DFSORT failed after processing about 6,956,799 . So if you know how many > > records you are going to process, then you can pass an estimated > filesize > > via DFSPARM like this > > > > //DFSPARM DD * > > OPTION FILSZ=E1000 > > /* > > > > Since DFSORT already processed about 6.9 million records, I just > estimated > > that you have 10 million records. You can change that number depending > on > > your input file. > > > > > > Further if you have any questions please let me know > > > > Thanks, > > Kolusu > > DFSORT Development > > IBM Corporation > > > > > > -- > > For IBM-MAIN subscribe / signoff / archive access instructions, > > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > > > > > Unless stated otherwise above: > IBM United Kingdom Limited - Registered in England and Wales with number > 741598. > Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 > 3AU > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
I'd ask over on MXG-L if there are methods for SAS to pass the record count and average record length estimate to DFSORT. Cheers, Martin Martin Packer Systems Investigator & Performance Troubleshooter, IBM +44-7802-245-584 email: martin_pac...@uk.ibm.com Twitter / Facebook IDs: MartinPacker Blog: https://mainframeperformancetopics.com Mainframe, Performance, Topics Podcast Series (With Marna Walle): https://anchor.fm/marna-walle Youtube channel: https://www.youtube.com/channel/UCu_65HaYgksbF6Q8SQ4oOvA From: "Gibney, Dave" To: IBM-MAIN@LISTSERV.UA.EDU Date: 19/01/2021 20:14 Subject: [EXTERNAL] Re: DFSORT and SAS Sent by:IBM Mainframe Discussion List Thank you. Adding the suggested parm allowed my job to succeed. Ultimately, there where 15,890,801 records. I looked again and there was not an ICE118I issued in the original job. I really don't know how, or what controls SAS (v9) passes to the HOST SORT Thanks again for the assistance. > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Sri h Kolusu > Sent: Tuesday, January 19, 2021 11:37 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > > ICE752I 0 FSZ=0 RE IGN=0 C AVG=188 0 WSP=0 E DYN=2313 16352 > > Dave, > > When DFSORT is invoked via program, it needs to know how many records it > is sorting so that it can optimally allocate resources needed to sort that > data. Since the program is feeding the records DFSORT does not have a clue > and the resources allocated are not enough to complete the sort and hence > the message ICE046A 0 SORT CAPACITY EXCEEDED - RECORD COUNT 6956799 > > Generally, DFSORT can automatically determine the input file size. However, > in a few cases, such as when an E15 supplies all of the input records or > when information about a tape data set is not available from a tape > management system, DFSORT cannot determine an accurate file size. > > Your joblog does not show one of the important message. You should have > gotten an ICE118I message > > ICE118I UNKNOWN FILE SIZE - RECOMMEND SPECIFYING FILSZ=EN TO > REDUCE RISK OF > FAILURE OR DEGRADED PERFORMANCE > > INVALID URI REMOVED > nter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.icem100/mb00083.htm__;!!JmPEg > BY0HMszNaDT!9_LGpASIz6lgFy3pDduAe2DC4jQhXQBtSHYnTd10a0weZdx0ie > 4mdHNVwnuilA$ > > DFSORT failed after processing about 6,956,799 . So if you know how many > records you are going to process, then you can pass an estimated filesize > via DFSPARM like this > > //DFSPARM DD * > OPTION FILSZ=E1000 > /* > > Since DFSORT already processed about 6.9 million records, I just estimated > that you have 10 million records. You can change that number depending on > your input file. > > > Further if you have any questions please let me know > > Thanks, > Kolusu > DFSORT Development > IBM Corporation > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
> Thank you. Adding the suggested parm allowed my job to succeed. Dave, Glad to hear that the job is successful. > Ultimately, there where 15,890,801 records. May be you should pass 18 million as estimated filesize to account for future growth. Thanks, Kolusu DFSORT Development IBM Corporation -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
Thank you. Adding the suggested parm allowed my job to succeed. Ultimately, there where 15,890,801 records. I looked again and there was not an ICE118I issued in the original job. I really don't know how, or what controls SAS (v9) passes to the HOST SORT Thanks again for the assistance. > -Original Message- > From: IBM Mainframe Discussion List On > Behalf Of Sri h Kolusu > Sent: Tuesday, January 19, 2021 11:37 AM > To: IBM-MAIN@LISTSERV.UA.EDU > Subject: Re: DFSORT and SAS > > > ICE752I 0 FSZ=0 RE IGN=0 C AVG=188 0 WSP=0 E DYN=2313 16352 > > Dave, > > When DFSORT is invoked via program, it needs to know how many records it > is sorting so that it can optimally allocate resources needed to sort that > data. Since the program is feeding the records DFSORT does not have a clue > and the resources allocated are not enough to complete the sort and hence > the message ICE046A 0 SORT CAPACITY EXCEEDED - RECORD COUNT 6956799 > > Generally, DFSORT can automatically determine the input file size. However, > in a few cases, such as when an E15 supplies all of the input records or > when information about a tape data set is not available from a tape > management system, DFSORT cannot determine an accurate file size. > > Your joblog does not show one of the important message. You should have > gotten an ICE118I message > > ICE118I UNKNOWN FILE SIZE - RECOMMEND SPECIFYING FILSZ=EN TO > REDUCE RISK OF > FAILURE OR DEGRADED PERFORMANCE > > https://urldefense.com/v3/__https://www.ibm.com/support/knowledgece > nter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.icem100/mb00083.htm__;!!JmPEg > BY0HMszNaDT!9_LGpASIz6lgFy3pDduAe2DC4jQhXQBtSHYnTd10a0weZdx0ie > 4mdHNVwnuilA$ > > DFSORT failed after processing about 6,956,799 . So if you know how many > records you are going to process, then you can pass an estimated filesize > via DFSPARM like this > > //DFSPARM DD * > OPTION FILSZ=E1000 > /* > > Since DFSORT already processed about 6.9 million records, I just estimated > that you have 10 million records. You can change that number depending on > your input file. > > > Further if you have any questions please let me know > > Thanks, > Kolusu > DFSORT Development > IBM Corporation > > > -- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
Re: DFSORT and SAS
> ICE752I 0 FSZ=0 RE IGN=0 C AVG=188 0 WSP=0 E DYN=2313 16352 Dave, When DFSORT is invoked via program, it needs to know how many records it is sorting so that it can optimally allocate resources needed to sort that data. Since the program is feeding the records DFSORT does not have a clue and the resources allocated are not enough to complete the sort and hence the message ICE046A 0 SORT CAPACITY EXCEEDED - RECORD COUNT 6956799 Generally, DFSORT can automatically determine the input file size. However, in a few cases, such as when an E15 supplies all of the input records or when information about a tape data set is not available from a tape management system, DFSORT cannot determine an accurate file size. Your joblog does not show one of the important message. You should have gotten an ICE118I message ICE118I UNKNOWN FILE SIZE - RECOMMEND SPECIFYING FILSZ=EN TO REDUCE RISK OF FAILURE OR DEGRADED PERFORMANCE https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.3.0/com.ibm.zos.v2r3.icem100/mb00083.htm DFSORT failed after processing about 6,956,799 . So if you know how many records you are going to process, then you can pass an estimated filesize via DFSPARM like this //DFSPARM DD * OPTION FILSZ=E1000 /* Since DFSORT already processed about 6.9 million records, I just estimated that you have 10 million records. You can change that number depending on your input file. Further if you have any questions please let me know Thanks, Kolusu DFSORT Development IBM Corporation -- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN