× The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.


  • Subject: Re: Logical file index size too small?
  • From: Pete Massiello <pmassiello@xxxxxxxxxxxxxxxx>
  • Date: Wed, 10 Dec 1997 15:05:42 -0500
  • Organization: OS Solutions International

Buck,

  You have found one of the wonders of OS/400 (Perhaps one of the best
things at that).  The *LF that you are examining is sharing an access
path with another file.  Most people use the word Access path and
Logical file interchangibly, which is really incorrect.  You can have
multiple logical files sharing an access path as long as the least
restrictive *LF is built first.  This is one of the features that our OS
Director product does.  It examines your LFs and determines where LFs
can share their access paths, and then makes OS/400 do it.  Did you
notice the field on your display that said Implicit Access Path
sharing.  We automate this whole process for you.

    In my opinion, you are better off spending more time figuring out
how to share those access paths, than to delete a LF that you really
need but you are deleting it for speed.  You get the speed and
accessibility by sharing of access paths. 

Buck Calabro wrote:
> 
> Hello again!
> 
> We've got a physical file with 2.7 million records and 10+
> logical files built over it.  In an attempt to speed up
> processing, we're analysing which logicals we can eliminate.
> While doing a DSPFD, I noticed that the index size for some
> of these logicals is really small: 16384.  This doesn't
> make any sense because the key length is 36 bytes: I'm
> expecting something closer to 95 megs!
> 
> Here's the section of DSPFD under Access Path statistics for
> logical file TRBRSQ08:
> 
> Index size  . . . . . . . . . . . . . . :             16384
> Access path valid . . . . . . . . . . . :            Yes
> Implicit access path sharing  . . . . . :            Yes
>   Access path journaled . . . . . . . . :            No
> File owning access path . . . . . . . . :     COMTST200/TRBRSQ08
> Member  . . . . . . . . . . . . . . . . :            TRBRSQ08
> Shared access path attributes
>   Maintenance . . . . . . . . . . . . . :            *IMMED
>   Access path recovery  . . . . . . . . :            *NO
>   Force keyed access path . . . . . . . :            *NO
>   Keys must be unique . . . . . . . . . :            No
> ...
> Based on file . . . . . . . . . . . . . . :            TRBPHY01
>   Library . . . . . . . . . . . . . . . . :            COMTST200
>   Member  . . . . . . . . . . . . . . . . :            TRBPHY01
>   Logical file format . . . . . . . . . . :            ZTRBRS08
>   Number of index entries . . . . . . . . :               2664901
> 
> How can this be right?  The index has 2.7 million records and
> is only 16k in size?  Another logical, TRBRSQ01 has a key length
> of 4, and it looks like:
> 
> Index size  . . . . . . . . . . . . . . :                    7933952
> Access path valid . . . . . . . . . . . :            Yes
> Implicit access path sharing  . . . . . :            No
> ...
> Based on file . . . . . . . . . . . . . . :            TRBPHY01
>   Library . . . . . . . . . . . . . . . . :            COMTST200
>   Member  . . . . . . . . . . . . . . . . :            TRBPHY01
>   Logical file format . . . . . . . . . . :            ZTRBRS01
>   Number of index entries . . . . . . . . :               2664901
>   Number of member accesses . . . . . . . :                     0
> 
> This looks much better: 2.7 million * 4 bytes = 10meg.  I've got
> 7.9 megs... close to what I expect.
> 
> I can see that RSQ08 is sharing it's access path, but it's being
> shared with itself!
> 
> So, my question is: what can make an access path size so much
> smaller than it should be?
> 
> Buck Calabro
> Commsoft


-- 
Pete Massiello
OS Solutions International 
Phone: (203)-744-7854  Ext 11.
http://www.os-solutions.com
mailto:pmassiello@os-solutions.com
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to "MIDRANGE-L@midrange.com".
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---


As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.