We are running SQL 2000 EI SP3 on a Windows 2000 Adv. Server.
We have recently setup a database for replication using transactional replic
ation every ten minutes.
Shortly after implementing this, we started receiving integrity errors on th
e database on a daily basis.
Many of the errors are requiring checktable with repair data loss to correct
the problem. The corruption has shown up on both replicated and non-replic
ated tables, but only occurs if replication is running. As soon as replicat
ion was stopped, the databa
se quit reporting integrity errors.
Please let me know if you need further information.
Has anyone else ran into this?I don't think replication has anything to do with it. Probably replication
is stressing your hard disk subsystem, leading to errors. So the real
culprit could be hardware. Check your event viewer and any hardware logs for
clues.
--
HTH,
Vyas, MVP (SQL Server)
http://vyaskn.tripod.com/
Is .NET important for a database professional?
http://vyaskn.tripod.com/poll.htm
"Ghost" <anonymous@.discussions.microsoft.com> wrote in message
news:B6B0118A-C15D-41A5-9E11-FA01F4AD14BD@.microsoft.com...
We are running SQL 2000 EI SP3 on a Windows 2000 Adv. Server.
We have recently setup a database for replication using transactional
replication every ten minutes.
Shortly after implementing this, we started receiving integrity errors on
the database on a daily basis.
Many of the errors are requiring checktable with repair data loss to correct
the problem. The corruption has shown up on both replicated and
non-replicated tables, but only occurs if replication is running. As soon
as replication was stopped, the database quit reporting integrity errors.
Please let me know if you need further information.
Has anyone else ran into this?|||Our first thought was hardware as well, however, a few things are pointing t
o something else causing the problem:
1. All of the hardware is brand new, and the server is running at about a q
uarter of our target performance. (This server is not yet hosting databases
that will be added later this year... During peak time we are seeing a max u
tilization of 15% across
all processors). The hardware vendor has already been over this server and
hasn't found anything, and there is no hardware related errors in the server
event logs
2. This is the first enterprise class server we have run in a SAN environme
nt. To eliminate the SAN hardware or network we moved all of the data to in
ternal drives on the server. The problem continues to occur.
3. This server is hosting multiple application databases. The only databas
e with errors is the replicated database. If this was hardware related we w
ould expect to see the other databases having problems, as they were using t
he same internal hardware
and SAN array to hold their data as the replicated database.
Any thoughts?|||Okay, could you post the exact errors you are seeing in the SQL Server error
logs?
Are you replicating text or image columns?
--
HTH,
Vyas, MVP (SQL Server)
http://vyaskn.tripod.com/
Is .NET important for a database professional?
http://vyaskn.tripod.com/poll.htm
"Ghost" <anonymous@.discussions.microsoft.com> wrote in message
news:34BB1D32-6935-40A8-B061-1372260A868B@.microsoft.com...
Our first thought was hardware as well, however, a few things are pointing
to something else causing the problem:
1. All of the hardware is brand new, and the server is running at about a
quarter of our target performance. (This server is not yet hosting databases
that will be added later this year... During peak time we are seeing a max
utilization of 15% across all processors). The hardware vendor has already
been over this server and hasn't found anything, and there is no hardware
related errors in the server event logs
2. This is the first enterprise class server we have run in a SAN
environment. To eliminate the SAN hardware or network we moved all of the
data to internal drives on the server. The problem continues to occur.
3. This server is hosting multiple application databases. The only
database with errors is the replicated database. If this was hardware
related we would expect to see the other databases having problems, as they
were using the same internal hardware and SAN array to hold their data as
the replicated database.
Any thoughts?|||Yes, There are both image and text columns in some of the tables.
Below is excerpt from the maintenance log on the errors we are getting:
[1] Database ServiceCenter: Check Data and Index Linkage...
[Microsoft SQL-DMO (ODBC SQLState: 42000)] Error 8929: [Microsoft][ODBC SQL Server Driv
er][SQL Server]Object ID 525244926: Errors found in text ID 462078345216 own
ed by data record identified by RID = (1:202550:3).
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Object ID 525244926. The te
xt, ntext, or image node at page (1:40468), slot 12, text ID 462078345216 is
referenced by page (1:202550), slot 3, but was not seen in the scan.
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Object ID 525244926. The te
xt, ntext, or image node at page (1:695828), slot 12, text ID 462078345216 i
s not referenced.
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Database 'ServiceCenter', i
ndex 'applicationm1.applicationm1_P' (ID 1589580701) (index ID 2). Extra or
invalid key for the keys:
[Microsoft][ODBC SQL Server Driver][SQL Server]Index row (1:81744:140) with values (app
lication = 'cc.find.problem' and label = 'build.s) points to the data row id
entified by (RID = (1:348126:8)).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation errors and 3
consistency errors in table 'problemm1' (object ID 525244926).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation errors and 1
consistency errors in table 'applicationm1' (object ID 1589580701).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation errors and 4
consistency errors in database 'ServiceCenter'.
[Microsoft][ODBC SQL Server Driver][SQL Server]repair_allow_data_loss is the minimum re
pair level for the errors found by DBCC CHECKDB (ServiceCenter ).|||There are a lot of bugs/problems, replicating text/image columns. First make
sure all participating servers are on the latest service pack. Then see if
you find any hits for those error numbers at: http://support.microsoft.com
And consider contacting Microsoft support.
--
HTH,
Vyas, MVP (SQL Server)
http://vyaskn.tripod.com/
Is .NET important for a database professional?
http://vyaskn.tripod.com/poll.htm
"Ghost" <anonymous@.discussions.microsoft.com> wrote in message
news:07926FA4-64B3-4038-8934-E2A3A631DDE3@.microsoft.com...
Yes, There are both image and text columns in some of the tables.
Below is excerpt from the maintenance log on the errors we are getting:
[1] Database ServiceCenter: Check Data and Index Linkage...
[Microsoft SQL-DMO (ODBC SQLState: 42000)] Error 8929: [Microsoft][ODBC SQL
Server Driver][SQL Server]Object ID 525244926: Errors found in text ID
462078345216 owned by data record identified by RID = (1:202550:3).
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Object ID
525244926. The text, ntext, or image node at page (1:40468), slot 12, text
ID 462078345216 is referenced by page (1:202550), slot 3, but was not seen
in the scan.
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Object ID
525244926. The text, ntext, or image node at page (1:695828), slot 12, text
ID 462078345216 is not referenced.
[Microsoft][ODBC SQL Server Driver][SQL Server]Table error: Database
'ServiceCenter', index 'applicationm1.applicationm1_P' (ID 1589580701)
(index ID 2). Extra or invalid key for the keys:
[Microsoft][ODBC SQL Server Driver][SQL Server]Index row (1:81744:140) with
values (application = 'cc.find.problem' and label = 'build.s) points to the
data row identified by (RID = (1:348126:8)).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation
errors and 3 consistency errors in table 'problemm1' (object ID 525244926).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation
errors and 1 consistency errors in table 'applicationm1' (object ID
1589580701).
[Microsoft][ODBC SQL Server Driver][SQL Server]CHECKDB found 0 allocation
errors and 4 consistency errors in database 'ServiceCenter'.
[Microsoft][ODBC SQL Server Driver][SQL Server]repair_allow_data_loss is the
minimum repair level for the errors found by DBCC CHECKDB (ServiceCenter ).
Monday, March 19, 2012
Data corruption on a replicated database
Labels:
adv,
corruption,
database,
microsoft,
mysql,
oracle,
replicated,
replication,
running,
server,
setup,
sp3,
sql,
transactional,
windows
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment