ASE 15.5, not able to select from a table:

[Former Member]

Hey,

Live backup restored to Reporting server and while selecting from the table ASE engine crashed
dbcc checktable(pm_schedule) says, page 8583904,partition ID 262552269 of object 262552269, contains an orphan row at offset 44 which is marked committed delete and is not.

An attempt was made to fetch logical page 8583905 for database pmt, object 'pm_schedule' (262552269), from cache 'default data cache'. Wrong logical page '1768843040' was found in cache.

1. select * into to different table failed
2. bcp failed
3. dbcc checktable((pm_schedule) - failed

Please do let me know if there is any option to recover the data completely.</p>

[Former Member]

Dears, As I conveyed before when I realized the issue took another backup and there was no issues with that backup. I am searching for an option to recover the table from the cache corrupted dump.

[javier_barthe]

It seems like a table corruption so start searching table on a clean backup or restore again a clean backup.

[Former Member]

looks like there might be cache corruption. Try restarting server. If the error is transient it will go away.

[Former Member]

I am fine with that option but just think about one thing. We all take backup believing the product does the integrity of the backup right? since we need to restore daily for the report generation we hit the issue immediately and could take another backup. If that was not the situation and we realize later where we will have that image???

[Former Member]

Dears,

It is a logical corruption happened during the dump, the source doesn't have any issue. I have tried all the index forcing combination it didn't help me.
dbcc checktable also doesnt work. Lookig for an option to recover the complete data

Table Corrupt: Attempted to get page 8583905, partition ID 262552269; got page 0, partition ID 0.

[former_member188958]

Is starting over by taking a fresh backup an option?

[victoria_normand]

Hi,

Prior of the BCP out or the select into as suggested by Cory, you might run dbcc tablealloc with fix option - dbcc tablealloc("pm_schedule", full, fix) - in single user mode. If the logical page number mentioned on the error comes from an index, drop index prior of BCP out (you may use ddlgen to have the complete SQL for recreate the indexes). 
Regards,
Victoria.

[Former Member]

Hello,

Try doing a select using another index if the object has any indexes.

select * from yourtable (index indexname)

select * from yourtable (index indexname)

or try writing "where clauses" that avoid the page.

select * from yourtable (index indexname) where col_a<abc or col_a>xyz

There are also dbcc  print page command...

You would want to examine the prior page in a table to see the next page pointer.

Cory Sane

If I answered your question correctly, please identify the reply as correct or helpful.