Re: Lisa Database Sizes
"David Young" <[email protected]> Mon, 20 Feb 2006 10:24:33 -0500
| Newsgroups | gmane.lisp.lisa.general |
|---|---|
| Message-ID | <[email protected]> |
------=_Part_12224_28401941.1140449073452 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Millions of concurrent objects for all of the Rete-based systems I've used, including KW, will be problematic. I think I've read that Forgy's "Rete III= " algorithm scales to this level, but I can't recall where I might have seen that. Roughly speaking, Lisa should be able to handle some "tens of thousands" of objects, given sufficient memory. If you want to manage millions of objects= , then it depends. If they all must be in memory concurrently, then you'll need perhaps a scheme for lazy evaluation; AllegroCache comes to mind, and KW can work with objects that live in relational databases. You'll sacrific= e some degree of speed with these approaches. The Loom system is another possibility, but I don't believe it will scale to millions of objects. Lisa wasn't originally written for very large knowledge bases, but I honestly don't know what the upper bound of data/rules might be before Lisa becomes sludgy. AllegroCache (www.franz.com) is actually the most interesting approach I've seen recent for integrating large data/knowledge bases into a Lisp app. You should visit the Franz site and read about it. Peace, david -----Original Message----- From: Brent Fulgham [mailto:[email protected]] Sent: Friday, February 17, 2006 4:45 PM To: Brent Fulgham Cc: Paolo Amoroso; [email protected] Subject: [Lisa-users] Lisa Database Sizes I have a few more questions that I don't find the answers to on the website= . 1. Knowledge Base Size: On thing I don't see covered on the website is a realistic feel for how big a database can get before it becomes unusable. Let's pretend I wanted to build a census system for all the hobbits in the shire. To continue the hobbit examples from the manual: If there are millions of hobbits, and I want to store dozens of facts about them, will Lisa be able to handle this much data? Or, more realistically, what if I have a few thousand entities, with a few dozen attributes on each? 2. Knowledge Base Persistence: Can the state of my Lisa session be made persistent in some way, or does it rely on saving a new lisp image to retain it? 3. Performance: I imagine that as the number of assertions and rules grow, the time require= d to find and trigger them increases. Is there a known upper bound for database size before performance issues become too annoying? Thanks, -Brent ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log file= s for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D1= 21642 _______________________________________________ Lisa-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/lisa-users This email message is for the sole use of the intended recipients(s) and ma= y contain confidential and privileged information of Bloodhound Software, Inc.. Any unauthorized review, use, disclosure is prohibited. If you are no= t the intended recipient, please contact the sender by reply email and destro= y all copies of the original message. This email message is for the sole use of the intended recipients(s) and ma= y contain confidential and privileged information of Bloodhound Software, Inc.. Any unauthorized review, use, disclosure is prohibited. If you are no= t the intended recipient, please contact the sender by reply email and destro= y all copies of the original message. -- For wisdom is more precious than rubies, and nothing you desire can compare with her. -- Proverbs 8:11 But all the world understands my language. -- Franz Joseph Haydn (1732-1809) ------=_Part_12224_28401941.1140449073452 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Millions of concurrent objects for all of the Rete-based systems I've used,= <br>including KW, will be problematic. I think I've read that Forgy's "= ;Rete III"<br>algorithm scales to this level, but I can't recall where= I might have seen <br>that.<br><br>Roughly speaking, Lisa should be able to handle some "= ;tens of thousands" of<br>objects, given sufficient memory. If you wan= t to manage millions of objects,<br>then it depends. If they all must be in= memory concurrently, then you'll <br>need perhaps a scheme for lazy evaluation; AllegroCache comes to mind, = and<br>KW can work with objects that live in relational databases. You'll s= acrifice<br>some degree of speed with these approaches. The Loom system is = another <br>possibility, but I don't believe it will scale to millions of objects.<= br><br>Lisa wasn't originally written for very large knowledge bases, but I= <br>honestly don't know what the upper bound of data/rules might be before = Lisa <br>becomes sludgy. AllegroCache (<a href=3D"http://www.franz.com">www.fran= z.com</a>) is actually the most<br>interesting approach I've seen recent fo= r integrating large data/knowledge<br>bases into a Lisp app. You should vis= it the Franz site and read about it. <br><br>Peace, david<br><br>-----Original Message-----<br>From: Brent Fulgh= am [mailto:<a href=3D"mailto:[email protected]">[email protected]</a>]<br>Sen= t: Friday, February 17, 2006 4:45 PM<br>To: Brent Fulgham<br>Cc: Paolo Amor= oso;=20 <a href=3D"mailto:[email protected]">[email protected]= forge.net</a><br>Subject: [Lisa-users] Lisa Database Sizes<br><br><br>I hav= e a few more questions that I don't find the answers to on the website.<br> <br><br>1. Knowledge Base Size:<br>On thing I don't see covered = on the website is a realistic feel for how big<br>a database can get before= it becomes unusable.<br><br>Let's pretend I wanted to build a census syste= m for all the hobbits in the <br>shire. To continue the hobbit examples from the manual:<br><= br>If there are millions of hobbits, and I want to store dozens of facts ab= out<br>them, will Lisa be able to handle this much data?<br><br>Or, more re= alistically, what if I have a few thousand entities, with a few <br>dozen attributes on each?<br><br>2. Knowledge Base Persisten= ce:<br>Can the state of my Lisa session be made persistent in some way, or = does it<br>rely on saving a new lisp image to retain it?<br><br>3. &nb= sp;Performance:<br> I imagine that as the number of assertions and rules grow, the time require= d<br>to find and trigger them increases. Is there a known upper = bound for<br>database size before performance issues become too annoying?<b= r><br> Thanks,<br><br>-Brent<br><br><br>------------------------------------------= -------------<br>This SF.net email is sponsored by: Splunk Inc. Do you grep= through log files<br>for problems? Stop! Download th= e new AJAX search engine that makes <br>searching your log files as easy as surfing the web. &n= bsp;DOWNLOAD SPLUNK!<br><a href=3D"http://sel.as-us.falkag.net/sel?cmd=3Dln= k&kid=3D103432&bid=3D230486&dat=3D121642">http://sel.as-us.falk= ag.net/sel?cmd=3Dlnk&kid=3D103432&bid=3D230486&dat=3D121642 </a><br>_______________________________________________<br>Lisa-users maili= ng list<br><a href=3D"mailto:[email protected]">Lisa-users@l= ists.sourceforge.net</a><br><a href=3D"https://lists.sourceforge.net/lists/= listinfo/lisa-users"> https://lists.sourceforge.net/lists/listinfo/lisa-users</a><br><br>This ema= il message is for the sole use of the intended recipients(s) and may<br>con= tain confidential and privileged information of Bloodhound Software,<br> Inc.. Any unauthorized review, use, disclosure is prohibited. If you are no= t<br>the intended recipient, please contact the sender by reply email and d= estroy<br>all copies of the original message.<br><br>This email message is = for the sole use of the intended recipients(s) and may contain confidential= and privileged information of Bloodhound Software, Inc.. Any unauthorized = review, use, disclosure is prohibited. If you are not the intended recipien= t, please contact the sender by reply email and destroy all copies of the o= riginal message. <br><br clear=3D"all"><br>-- <br>For wisdom is more precious than rubies,<b= r>and nothing you desire can compare with her.<br> -- Proverbs 8= :11<br><br>But all the world understands my language.<br> -- Fra= nz Joseph Haydn (1732-1809)=20 ------=_Part_12224_28401941.1140449073452-- ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642