Help! Index corrupting? DBCC CheckTable failing?

We are having quite a time since moving a large database to a new
server (actually built new server, renamed as old to make seamless for
users, etc.)

Import 104 million row database (5 column) into table (CD_Assets_bad2)
from existing (CD_Assets):
Account(varchar(8))
TransactionDate(datetime(8)
Flow(varchar(1))
Category(varchar(7))
TotalValue(decimal(8))

Run DBCC CheckTable - no issues.
Create 4 non clustered indexes (3 single column, 1 two-column). All
indexes create fine.

Run DBCC CheckTable again and receive the following:
Server: Msg 8951, Level 16, State 1, Line 1

Table error: Table 'CD_Assets_bad2' (ID 244195920). Missing or invalid
key in index 'idx_totalvalue' (ID 7) for the row:

Server: Msg 8955, Level 16, State 1, Line 1

Data row (1:11154499:98) identified by (RID = (1:11154499:98) ) has
index values (TotalValue = -10).

Server: Msg 8952, Level 16, State 1, Line 1

Table error: Database 'CD', index 'CD_Assets_bad2.idx_totalvalue' (ID
244195920) (index ID 7). Extra or invalid key for the keys:

Server: Msg 8956, Level 16, State 1, Line 1

Index row (1:20855652:338) with values (TotalValue = -0

4) points to the data row identified by (RID = (1:11154499:98)).

DBCC results for 'CD_Assets_bad2'.

There are 104397173 rows in 677904 pages for object 'CD_Assets_bad2'.

CHECKTABLE found 0 allocation errors and 2 consistency errors in table
'CD_Assets_bad2' (object ID 244195920).

repair_fast is the minimum repair level for the errors found by DBCC
CHECKTABLE (CD.dbo.CD_Assets_bad2 ).

Any ideas? It seems like some sort of corruption, but the index
creates fine. If anyone can help please let me know. If I can provide
any addtional information that might help, please let me know.

Thanks,
David
david.schwartz@.schwab.com[posted and mailed, please reply in news]

David Schwartz (david.schwartz@.schwab.com) writes:
> Any ideas? It seems like some sort of corruption, but the index
> creates fine. If anyone can help please let me know. If I can provide
> any addtional information that might help, please let me know.

Both errors 8252 and 8956 have articles in Books Online. (Just search
for the numbers.)

Both articles says that the index must be reparied or dropped.

Since your indexes appears to be corrupted when you build them, it smells
a hardware problem to me. I would probably open a case with Microsoft
to investigate this.

--
Erland Sommarskog, SQL Server MVP, sommar@.algonet.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techin.../2000/books.asp|||Try Creating clusterd index and import the data after dropping all the
existing indexes in the destination table.

david.schwartz@.schwab.com (David Schwartz) wrote in message news:<32d60003.0312021706.10bee3d6@.posting.google.com>...
> We are having quite a time since moving a large database to a new
> server (actually built new server, renamed as old to make seamless for
> users, etc.)
> Import 104 million row database (5 column) into table (CD_Assets_bad2)
> from existing (CD_Assets):
> Account(varchar(8))
> TransactionDate(datetime(8)
> Flow(varchar(1))
> Category(varchar(7))
> TotalValue(decimal(8))
> Run DBCC CheckTable - no issues.
> Create 4 non clustered indexes (3 single column, 1 two-column). All
> indexes create fine.
> Run DBCC CheckTable again and receive the following:
> Server: Msg 8951, Level 16, State 1, Line 1
> Table error: Table 'CD_Assets_bad2' (ID 244195920). Missing or invalid
> key in index 'idx_totalvalue' (ID 7) for the row:
> Server: Msg 8955, Level 16, State 1, Line 1
> Data row (1:11154499:98) identified by (RID = (1:11154499:98) ) has
> index values (TotalValue = -10).
> Server: Msg 8952, Level 16, State 1, Line 1
> Table error: Database 'CD', index 'CD_Assets_bad2.idx_totalvalue' (ID
> 244195920) (index ID 7). Extra or invalid key for the keys:
> Server: Msg 8956, Level 16, State 1, Line 1
> Index row (1:20855652:338) with values (TotalValue = -0
> 4) points to the data row identified by (RID = (1:11154499:98)).
> DBCC results for 'CD_Assets_bad2'.
> There are 104397173 rows in 677904 pages for object 'CD_Assets_bad2'.
> CHECKTABLE found 0 allocation errors and 2 consistency errors in table
> 'CD_Assets_bad2' (object ID 244195920).
> repair_fast is the minimum repair level for the errors found by DBCC
> CHECKTABLE (CD.dbo.CD_Assets_bad2 ).
> Any ideas? It seems like some sort of corruption, but the index
> creates fine. If anyone can help please let me know. If I can provide
> any addtional information that might help, please let me know.
> Thanks,
> David
> david.schwartz@.schwab.com

Help! Create index with substring

Good day!

We had decided to migrate from Oracle to SQL Server, so faced some problems. Using Oracle we could create indexes like that

create index obj_id_cnum on obj_id (substr(cnum,1,2))

But Microsoft SQL Server doesn't allow this code. How can we do the same using SQL Server. Thanks you.

dynamic sql...

declare @.SQL nvarchar(100)

set @.SQL = 'create index [obj_id_cnum] on object_id(' + substring(cnum,1,2) + ')'

exec sp_executesql @.SQL

|||

Is obj_id a table, I am assuming? If so, then you are right, you cannot apply an index on a function directly. Two ways you can approach this:

1. If using enterprise edition, you can index a view and it will be used:

create view obj_id_indexed
as
select obj_id_key, cnum, substring(cnum,1,2) as cnumSub,
from obj_id

create unique clustered index on obj_id_cnum(obj_id_key, cnumSub)

Now your queries will see this index on the the view and apply it just like the

2. You can however add a computed column to your table and then index it:

alter table obj_id
add cnumSub as substring(cnum,1,2)

I would suggest that all of these techniques are probably the "wrong" way to go about this sort of thing. Almost any time you need to use a substring on a value in a SQL table there is a problem with normalization. Better would be to break the column into two parts, and then you have a much better chance of making the indexes work for you.

|||Thanks you very much! I do appreciate your help!|||

Derek Comingore - RSC wrote:

dynamic sql...

declare @.SQL nvarchar(100)

set @.SQL = 'create index [obj_id_cnum] on object_id(' + substring(cnum,1,2) + ')'

exec sp_executesql @.SQL

set @.SQL = 'create index [obj_id_cnum] on object_id(' + substring(cnum,1,2) + ')' gives an error 'Invalid column name 'cnum'.

|||Note that we are considering adding function / expression based indexes for a future version of SQL Server. For now, the easiest workaround is to use an indexed computed column.