December 11, 2001 at 9:49 am
NT Error:
I/O error (bad page ID) detected during read of BUF pointer = 0x11e92fc0, page ptr = 0x2404e000, pageid = (0x1:0x9dda), dbid = 10, status = 0x801, file = E:\MSSQL7\data\FCP_INNOVA_HISTORY_Data.MDF.
Followed the instruction from BOL and repaired the database with DBCC. Had to resort to using option "REPAIR_ALLOW_DATA_LOSS".
My question to the board is this: I've been administering SQL Sever databases for about a year now. This is the first 823 that I have hit. I need to report to management on this issue. Is it normal or expected to get an 823 once or twice a year? Do I need to plan for more of these items. Is this indicitive of more errors to come?
Any advice / experience with 823 error would be appreciated.
Cheers.
- Brendan
December 11, 2001 at 9:54 am
Is this NT or 2K, also what version of SQL.
In working with SQL 7 for 2 years (3 servers) and SQL 2000 for a year (7 servers), never had an 823. Did have some in SQL 6.5 (not same error, but similar).
Steve Jones
December 11, 2001 at 10:01 am
SQL Server 7.0 Service Pack 2 running on windows 2000.
Five databases. Four databases under 500 MB. The database with the 823 is just under 4 GB.
December 11, 2001 at 10:10 am
A quick search on Technet for "sql 823" gives these articles. Any apply?
http://support.microsoft.com/default.aspx?scid=kb;EN-US;q274310
http://support.microsoft.com/default.aspx?scid=kb;EN-US;q281809
Steve Jones
December 12, 2001 at 7:10 am
Bad news....got another 823 last night.
Steve - thanks for the links to the two MS articles. I do not believe I am hitting either of them. I will be calling Premiere Support this morning and get Microsoft on this item.
Good news, we do some good logging and know the two statements that caused the error. One statement was an "select insert into temp table", the other statement was "insert into temp table". Common thread is a temporary table.
Monday nights statement that caused the 823:
/* Get complete batches */
SELECT i.reject_reason reject_reason,
pp.object_id pp_oid,
pp.owner_object_id uow_oid,
pp.value_object_id item_oid,
LEFT( i.document_id, 3) batch_nbr,
LEFT( uw.activity_name, 10 ) Activity
INTO #tmp_cgf_completed_batches
FROM persistent_property pp,
unit_of_work uw,
item i
WHERE ( pp.owner_object_id = uw.object_id )
AND ( uw.work_type = "batch" )
AND ( value_object_id = i.object_id )
AND ( uw.activity_name = 'Complete' AND uw.status_id = 0 )
AND ( i.reject_reason = 0 )
AND ( i.fcp_trade_dt = @ldtm_trade_dt)
ORDER BY batch_nbr
Tuesday nights statement that cuased the 823:
INSERT INTO #tmp_deposit_slip (object_id, owner_object_id, primitive_value)
SELECT object_id, owner_object_id, primitive_value
FROM persistent_property
WHERE name = 'accountNum'
AND fcp_trade_dt = @ldtm_trade_dt
===================================================
I'll let the board know how I make out with Microsoft.
- Brendan
December 12, 2001 at 10:02 am
PLs do. This is one reason I avoid Temp tables if possible. There are more bugs associated with temp tables.
Steve Jones
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply