Re: oncheck -cD errors
Thomas J. Girsch saw oncheck -cD errors on nearly every record of a multi-million row table. Art Kagel said it was not a real problem since oncheck repaired the damage, possibly from bitmap pages not flushed after a crash, and noted one bitmap page can cover over 10 million rows. bozon asked whether two instances shared the same space. Girsch had heard it was harmless unless it persists.
Auto-generated by Claude from the posts below — may be imperfect; read the full thread.
On Aug 26, 11:36 pm, "Thomas J. Girsch" <tgir...@NOSPAM.gmail.com>
wrote:
> All:
>
> I'm seeing bazillions of these errors on some tables when I run oncheck -cD:
>
> WARNING: data page 0xe495 in tablespace 0x4f00002 appears to be
>
> more or less full than is indicated in the bitmap.
>
> Bitmap mode: 0xc, Calculated mode: 0x4.
>
> WARNING: The bitmap mode for page 0xe495 has been reset from 0xc to 0x4.
>
> Anything to worry about?
>
> Thanks,
>
> - TJG
Not a real problem. The oncheck repaired the damage. Could be that a
crash or non-standard shutdown in the past caused modified bitmap
pages to not be flushed to disk.
Art S. Kagel
Art S. Kagel wrote:
>
> Not a real problem. The oncheck repaired the damage. Could be that a
> crash or non-standard shutdown in the past caused modified bitmap
> pages to not be flushed to disk.
>
> Art S. Kagel
>
I can't imagine that would be it. It occurred for nearly every record
in a table with millions of rows. But I've heard other reports that
it's nothing to worry about unless it persists.
Thanks,
- TJG
On Aug 28, 12:20 am, "Thomas J. Girsch" <tgir...@NOSPAM.gmail.com>
wrote:
> Art S. Kagel wrote:
>
> > Not a real problem. The oncheck repaired the damage. Could be that a
> > crash or non-standard shutdown in the past caused modified bitmap
> > pages to not be flushed to disk.
>
> > Art S. Kagel
>
> I can't imagine that would be it. It occurred for nearly every record
> in a table with millions of rows. But I've heard other reports that
> it's nothing to worry about unless it persists.
>
> Thanks,
>
> - TJG
You don't have 2 instances sharing the same space do you? Otherwise I
agree with what you have heard.
On Aug 28, 12:20 am, "Thomas J. Girsch" <tgir...@NOSPAM.gmail.com>
wrote:
> Art S. Kagel wrote:
>
> > Not a real problem. The oncheck repaired the damage. Could be that a
> > crash or non-standard shutdown in the past caused modified bitmap
> > pages to not be flushed to disk.
>
> > Art S. Kagel
>
> I can't imagine that would be it. It occurred for nearly every record
> in a table with millions of rows. But I've heard other reports that
> it's nothing to worry about unless it persists.
But each bitmap page describes over 4000 pages and depending on
pagesize and rowsize a single bitmap page can affect more than
10million rows!
Art S. Kagel