Sign up on the Revelation Software website to have access to the most current content, and to be able to ask questions and get answers from the Revelation community

At 28 JUL 1999 12:47:24PM Jim Horvath wrote:

Every now and then someone notices a random discrepancy on a report generated by our AREV application. The fix is almost always to rebuild the indexes of the associated tables. It's not that the indexes have GFE's, or that the dedicated indexer on our system is not running. It just appears that some records were never indexed, and never will be without a complete rebuild.

My question is, short of getting a GFE, how do you know that an index file is out-of-sync with the data? (i.e. VERIFYLH doesn't show any errors). What would cause the indexer to miss random records, but otherwise appear to be operating normally?

Jim


At 28 JUL 1999 02:56PM [email protected] - [url=http://www.sprezzatura.com]Sprezzatura Group[/url] wrote:

If the write aborts before finishing, it could happen. Same thing with the

dedicated indexer crashing. If locking isn't occuring, it could be that a

station is overwriting the transaction files.

[email protected]

Sprezzatura Group

www.sprezzatura.com_zz.jpg


At 29 JUL 1999 01:51PM Jim Horvath wrote:

] If the write aborts before finishing, it could happen. Same thing with the

] dedicated indexer crashing. If locking isn't occuring, it could be that a

] station is overwriting the transaction files.

This answers my second question, but not the first, - is there a systematic way to verify index validity - that each index points to the matching record and that every record is indexed?

If there were locking problems, wouldn't I be seeing GFE's? (I'm not)

TIA

Jim


At 29 JUL 1999 03:42PM [email protected] onmouseover=window.status=why not click here to send me email?;return(true)", [url=http://www.sprezzatura.com" onMouseOver=window.status=Why not click here to visit our web site?';return(true)]Sprezzatura Group[/url] wrote:

Well GFEs do not always occur if locking doesn't work. The only way to systematically verify indexes is ultimately almost the same as building them!

[email protected]

Sprezzatura Group

World Leaders in all things RevSoft


At 29 JUL 1999 04:21PM [email protected] - [url=http://www.sprezzatura.com]Sprezzatura Group[/url] wrote:

You don't really, unless you write a program to do validation. While we have programs that validate index definitions and ensure matching dictionary references and pointers, we don't have a program that validates the actual indexed keys and data. Perhaps someone else does.

[email protected]

Sprezzatura Group

www.sprezzatura.com_zz.jpg

View this thread on the forum...

  • third_party_content/community/commentary/forums_nonworks/d8d1ba50bcba8cfe852567bc005c3b52.txt
  • Last modified: 2023/12/28 07:40
  • by 127.0.0.1