Hi, I now understand what you are saying.
I guess that when it reads for example 3000 records, it just drops record 1-2000 and uses the trailing 1000 records. So for you, this results in a massive number of SQL-searches instead of reading through the table once. In this case, I guess a direct SQL call would be only workable solution... I think that the best way for BMC to handle/implement this would be to use the chunk-size you specify for the table-field, with a possible maximum-limit equal to the system wide max-entries-returned-by-get-list... I must confess that I typically work with fewer records in filter-tables. Best Regards - Misi, RRR AB, http://www.rrr.se Products from RRR Scandinavia: * RRR|License - Not enough Remedy licenses? Save money by optimizing. * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. Find these products, and many free tools and utilities, at http://rrr.se. > That's the thing....the table itself isn't set to chunk....the chunking is > happening at the server and affects ALL tables the server needs to read. > The table in question has 4 columns, all that need to at some point have a > colsum done on them. Based on my reading of the server side table > chunking, > it's designed to keep the memory usage down by handling x records at a > time...but in this case that's not what it does....it simply causes the > server to increase the chunk size by that amount until it's dealing with > the > entire table....and it does this automatically when you ask for a > colsum....it knows that all you want is a sum of the values, yet it does > massively more work than is necessary to get the information and ends up > taking massively more time than necessary....thus I was looking for Doug's > view on the subject as he seems to always have an eloquent response to > such > design questions. > > -----Original Message----- > From: Action Request System discussion list(ARSList) > [mailto:arsl...@arslist.org] On Behalf Of Misi Mladoniczky > Sent: Monday, April 19, 2010 11:11 PM > To: arslist@ARSLIST.ORG > Subject: Re: COLSUM > > Hi, > > I seems to be doing what you asked it to do. > > The way both client and FLTR tables work, is by filling the table with > data (in memory). > > If you tell it to chunk, it apparently reads it that way. > > Create a new table-field without chinking, with a single column, and do > the COLSUM there instead. > > Best Regards - Misi, RRR AB, http://www.rrr.se > > Products from RRR Scandinavia: > * RRR|License - Not enough Remedy licenses? Save money by optimizing. > * RRR|Log - Performance issues or elusive bugs? Analyze your Remedy logs. > Find these products, and many free tools and utilities, at http://rrr.se. > >> Doug, >> Since you have been replying to list posts more frequently, I figured I >> would ask you a question real quick. >> >> We have several pieces of workflow that use filter side tables and >> COLSUM. >> In situations where there are more records in the table than the 'Server >> Table Field Chunk Size', I have noticed that it does a 'select first x' >> on >> the table columns, and it does it again till it finds a number higher >> than >> the number of records returned. In our situation we had the setting at >> 1000 >> and had over 50K records in the table...so the SQL log shows select >> first >> 1000, then 2000, 3000, 4000, etc till it hits a value above the count in >> the >> table. This of course kills our performance when we are just doing a >> sum >> on >> a column....so I converted them to direct SQL doing SUM() and aren't >> having >> problems anymore. Why was this done as a method to perform >> COLSUM?....it >> doesn't seem efficient to me at all. >> >> > ____________________________________________________________________________ > ___ >> UNSUBSCRIBE or access ARSlist Archives at www.arslist.org >> attend wwrug10 www.wwrug.com ARSlist: "Where the Answers Are" >> >> -- >> This message was scanned by ESVA and is believed to be clean. >> > > ____________________________________________________________________________ > ___ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug10 www.wwrug.com ARSlist: "Where the Answers Are" > > _______________________________________________________________________________ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > attend wwrug10 www.wwrug.com ARSlist: "Where the Answers Are" > > -- > This message was scanned by ESVA and is believed to be clean. > _______________________________________________________________________________ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org attend wwrug10 www.wwrug.com ARSlist: "Where the Answers Are"