You ought to run SET EMSG ON so that you (and others in particular) can see which module is issuing a message. From your trace below, I'd say it takes you seven seconds to read the files and do whatever else you do to massage them. Time between the end of the PIPE command and the R; message is likely CMS storage management cleaning up.
j. On 2 March 2010 20:19, Alan Winson <[email protected]> wrote: > Thank you, John, for the NXPIPE (version 1.0112). Tested it just now and > it > does cause the pipe to stop fairly quickly with a non-zero return code, > which is just what I wanted. > > (It doesn't seem to terminate instantly, but that is probably because of > failure processing that I don't fully understand.) > > For my test case I lowered the amount of main storage so it would run out > quickly. Here is an edited console log showing the failure: > > 14:01:33 Loading newer version of pipes > 14:01:33 1.0112 2 Mar 2010 > 14:01:33 11 Rev 0C Mod 0001 > 14:01:33 110C0001 > 14:01:33 Number of files selected for reading: 22 > 14:01:33 Files contain 1,355,558 records. > 14:01:33 Reading CUSIPCLK files, writing CUSIP output. > 14:01:40 PIPSTO122E Insufficient free storage. > 14:01:40 PIPMSG004I ... Issued from stage 5 of pipeline 1 name "huge". > 14:01:40 PIPMSG001I ... Running "lookup count autoadd w1". > 14:01:40 CUSIPCLK 20100226 SPS:SPDSIDC2.CUSPBILL. F 1024 62311 > 15578 2010-03-01 04:30:54 > 14:01:40 PIPDBK1436E FIXED specified, but no record length specified and no > input. > 14:01:40 PIPMSG004I ... Issued from stage 4 of pipeline 1 name "GOODPACK". > 14:01:40 PIPMSG001I ... Running "> CUSIPCMB SM100228 VMCMGC:BPWINSO. > fixed". > 14:01:40 394 *-* 'PIPE (name huge end ? stoperror)', /* abort if > insufficient storage > 14:01:40 */ ' stem input_files.', /* fn ft (fm|dir) etc. */ > '| eofback con > 14:01:40 sole', /* show details */ '| getfiles', > /* all > 14:01:40 records from all files */ , /* concatenate in case take_stage > is > a find or nf > 14:01:40 ind: */ take_stage ||, /* take only a sample > in > dbg mode */ > 14:01:40 '| spec w1 1 w2 nw /1/ nw w3-* nw', /* insert day count 1 as w3 > */ , > 14:01:40 /* and reduce no. bytes/record */ '| > acc: > lookup count > 14:01:40 autoadd w1', /* don't care about count, but */ , > 14:01:40 /* forces dump of masters at end */ '| copy', > /* f > 14:01:40 ree lookup after match */ '| spec t: w2 . d: w3 .', /* > detail appearanc > 14:01:40 e/day counts */ 'set #0:=t', /* detail > appearance count */ > 14:01:40 'set #1:=d', /* detail day count */ > 'read t: w2 . d: > 14:01:40 w3 .', /* master appearance/day counts */ 'set #0+=t', > 14:01:40 /* accumulate appearance count */ 'set #1+=d', > /* accu > 14:01:40 mulate day count */ 'w1 1', /* CUSIP */ > 'print # > 14:01:40 0 strip nw', /* updated appearance count */ 'print > #1 > strip nw', > 14:01:40 /* updated day count */ 'w4-* nw', > /* flags */ > 14:01:40 '| w1: fanout', /* artifice for label ordering > */ > '? acc:', > 14:01:40 /* no secondary input or output */ '? > w1:', '| w2: no > 14:01:40 t fanout', /* feed 2nd (lookup 4th) first */ '| > acc:', > 14:01:40 /* tertiary input, add to master */ '| not chop > 10', > 14:01:40 /* 3rd o/p, gr totals; drop count */ , > /* c > 14:01:40 usip, appct, dayct, <flags> */ out_stages, > /* 1 or 2 sets > 14:01:40 of 1st 3 files */ '? w2:', /* fanout > tertiary */ '| > 14:01:40 acc:', /* 4th input, delete from master */ > '? var rexx > 14:01:40 pgm', /* accum flag1 value totals */ '| f1:' > 14:01:40 /* to 2nd f1 input in out_stages */ > 14:01:40 +++ RC(-4094) +++ > 14:01:40 Error -4094 in line 394: > 14:01:40 Source: 'PIPE (name huge end ? stoperror)', /* abort if > insufficient storage */ > 14:01:40 Command: PIPE (name huge end ? stoperror) stem input_files. | > eofback console | getfiles | spec w1 1 w2 nw /1/ nw w3-* nw > | acc: lookup count autoadd w1 | copy | spec t: w2 . d: w3 . set #0:=t set > #1:=d read t: w2 . d: w3 . set #0+=t set #1+=d w1 1 prin > t #0 strip nw print #1 strip nw w4- > 14:01:40 Exiting with code -4094 >
