I have a strange problem, at least it seems strange to me.
However, if i run this script on my client it runs through perfect:
FST_DATEI_FZH_SERVICEFALL_EVENT_PROGR1: load Text(FGNR7) as FGNR7_FST_DATEI_PROGR, Text(FGNR17) as FGNR17_FST_DATEI_PROGR, timestamp(DIAG_DATETIME) as DIAG_DATETIME_FST_DATEI_PROGR, time(DIAG_DATETIME - (10/86400)) as DIAG_DATETIME_FST_DATEI_PROGR_UG, time(DIAG_DATETIME + (10/86400)) as DIAG_DATETIME_FST_DATEI_PROGR_OG, timestamp(ERRTIME) as ERRTIME_FST_DATEI_PROGR, Year(ERRTIME)& '-' &num(Month(ERRTIME),00) as ERRTIME_JAHR_MONAT_FST_DATEI_PROGR, DATEINAME as DATEINAME_FST_DATEI_PROGR, 1 as ZAEHL_PROGR_ABG_FST_DATEI_PROGR from $(vPfad)FST_DATEI_PROGR.qvd (qvd) where exists (FGNR17_DIM_FAHRZEUG, FGNR17) ;
FZH_SERVICEFALL_EVENT_PROGR: load Text(FGNR7_FZA) as FGNR7_FST_DATEI_PROGR, Text(FGNR17_FZA) as FGNR17_FST_DATEI_PROGR, Time(Frac(timestamp(REFERENZ_DT))) as REFERENZ_DT_FZH_SERVICEFALL_EVENT_TYP_PROGR_ABG, num(EVENT_ID_EVENT) as EVENT_ID_FZH_SERVICEFALL_EVENT_TYP_PROGR_ABG, num(EVENT_ID_PROGR) as EVENT_ID_FZH_EV_PROGRAMMIERUNG_ABG, STATUS as STATUS_FZH_EV_PROGRAMMIERUNG_ABG, if (isnull(EVENT_ID_EVENT),0,1) as ZAEHL_PROGR_ABG_FZH_SERVICEFALL_EVENT_TYP_PROGR_ABG_TEMP, if (isnull(EVENT_ID_PROGR),0,1) as ZAEHL_PROGR_ABG_FZH_EV_PROGRAMMIERUNG_ABG_TEMP, timestamp#(mid(timestamp(INSERT_DATE_FZA,'DD.MM.YYYY hh:mm:ss'),1,16),'DD.MM.YYYY hh:mm') as INSERT_DATE_FZH_EV_PROGRAMMIERUNG_ABG from $(vPfad)FZH_SERVICEFALL_EVENT_PROGR.qvd (qvd) ;
I don't have a reply to this, I'm just looking for a solution or explanation for this error too:
Interval Fields List not matching number of fields.
What is that? This suddenly started appearing in a QV document I'm running every month, made no changes whatsoever. I even added floor() functions to interval and matching field dates, so that there are no format integrity issues.
I couldn't provide an answer only some hints. Within the help to the intervalmatch are remarks regarding the included data - of course the interval-fields must be numeric but also that there could be unexpected results if there are NULL's which should be checked/replaced in beforehand.
Therefore I suggest to make a quick test with a smaller subset of the data - with any where-conditions which surely excludes possible NULL's and if this worked to extend the dataset to find the causing records (maybe there is an issue with the data-quality) and/or to implement logics to replace the NULL's with real values.
Beside this you could consider to replace the inbuilt intervalmatch-feature with an internal while-loop like:
load [from]+ iterno() - 1 as Date, * from Source while [from] + iterno() - 1 <= to;