A customer of ours has reported a intermittent error using one of our applications to receive some stock.
Under normal circumstances the application updates about 10 tables in a single transaction, inserting to some, updating others. After the stock receipt they print off a Crystal report against the receipt data.
The customer is saying that (very) occasionally they perform a receipt (no errors) and print off the report (no problems) yet when they come to access the data some time later it is almost as if no receipt has been made.
I have seen their data and can see what they mean. None of the expected data updates seem to have taken place. And yet there is the report sat there, indicating that it must have...
Looking closely at the data I can see breaks in the sequencing in certain identity column-carrying tables, during the period when the stock receipt was made. To all intents and purposes, therefore, it looks as though a rollback has occurred.
Is this possible? Could a Crystal report show uncommitted data which is then rolled back? How 'long' can a rollback take? Can it be initiated in some other way?
NB: We have done all the obvious things like:
- checking they have received against the correct DB;
- checking that no app or procedure can remove data in this way.
- verified that the report is reporting from the correct place.
etc.
The 'missing' data is so perfectly removed, my instinct says it must be a rollback but I can't see how this can be (yet).
All suggestions gratefully received - how could I track down whether this was occurring? Or is it my fevered imagination?
Many thanks!!!!
pmb
ps: According to our records, this customer is running SQL 7 SP3.
I am wondering if there is a correlation between the size of a database and the amount of RAM required in the server. Here's what's going on:
NT4.0 SP 6a, SQL 7.0 2-800 MHz Pentium III Processors 512 MB RAM Swap Files: two, divided evenly between two partitions, min 512 per partition (dynamic) 3- 9.2 GB Drives, RAID 5
SQL Database size: 2.5 GB
The SQL administrators have a scheduled job set to dump the database every four hours, so they have 6 of these dump files per day. Seagate Backup Exec v 6.? backs up those files every evening. Of course, this is a mission-critical application.
As the NT/Hardware administrator, I was tasked to flash upgrade all the hardware that could be flashed. That happened about three weeks ago. Ever since I flashed it, getting the data dumps has been an inconsistent process. Sometimes it works, sometimes it errors out. The error message is " Lost Delayed - Write Data: The system was attempting to transfer file data from buffers to (path). The write operation failed, and only some to the data may have been written to the file."
On and off during the day, you can't even copy a small file on the server without receiving the same message.
Before I flashed the server, it was blue-screening and needing to be rebooted about once every two weeks. Since I flashed it, the server hasn't been rebooted, because it hasn't blue-screened.
The database dumps are about 2 GB. We were thinking that the way the dump process worked was that a snapshot of the data was put into memory before it was written to disk. We thought since the dump size was larger than Physical Memory + Virtual Memory, that that could be causing the error.
What do you guys think? 1. Is dumping the database first the preferred way to get a backup, or should a backup directly to tape be scheduled instead? 2. Do we have enough RAM for the database? 3. Are there special settings we should set for the RAID controller card to maximize throughput for SQL?
Thanks again. I am definitely not a SQL DBA and any information you can provide would be really helpful.
Hello experts, I made a huge mistake on a production server, ran an update without the WHERE clause. UPDATE myTableSET Field1 = 'SOmetext',Field2 = 'sometext2'Is there a way to rollback? I read something a while ago about transaction logs.SOS!Thanks
HI, Suppose I execute any transaction say inserted some rows. I want to delete my previous transaction using rollback. What is the proper syntax for that. In Query Window: Insert ---------- executed rollback. it was asking for name of the transaction. How to give that. Help me out Thanks Bye
This is my stored procedure. Problem is that a rollback does not happen when there is an error ? How can this be, am starting and ending the transaction?
I have rollback on one of the users. I tried killing her process through Enterprise Manager and Query Analyzer, but I still can see her process on the screen. However, her login is grayed out after I execute the kill query.
I have encountered this problem several times, and the only way I can get clear the rollback is by rebooting the server.
Does anyone have a sugeestion on how to clear the rollback without rebooting.
Hi All A question on DTS ROLLBACK CAPABILITY. Is it possible to setup a DTS pkg to roll back the loaded files if a step fails? If a step fail records inserted by Data Pump to a table should be removed. Is this possible? If so how can we achieve this?
is it possible to roll back a transaction here? i heard that it will be possible only if we wrote the code only in certain block.please tell me the block.
The code inserts into table TABLES. TABLES has a primary key. If the insert statement has no error it should insert 'ROWS INSERTED' to table CONTROL and update table DATETRACKING else insert 'ROWS NOT INSERTED' to table CONTROL and do not update table DATETRACKING. 1- I will like to improve my code if it is possible. 2- Can I use a rollback?
DECLARE @ROWCOUNT INT DECLARE @ERROR INT DECLARE @LASTDT DATETIME
SET @LASTDT =GETDATE() BEGIN TRAN insert tables select 4
SELECT @ROWCOUNT = @@ROWCOUNT, @ERROR = @@ERROR IF (@ERROR <> 0) BEGIN INSERT CONTROL SELECT 'ROWS NOT INSERTED', @ROWCOUNT END ELSE BEGIN INSERT CONTROL SELECT 'ROWS INSERTED', @ROWCOUNT END COMMIT TRAN
IF (@ROWCOUNT > 0) BEGIN UPDATE DATETRACKING SET LASTDATEDL = GETDATE() END
I have SQL server 2000 with SP4, 1.Some of the reason,I want roll back SP4 and I want replace SP3. 2.Is it true SP4 slow down the Quey performance compare with SP3.
My rollback does not work In my SP I want any of cmdS,cmdS2,cmdS3,cmdS4 produces error Rollback must be executed for all of cmdS,cmdS2,cmdS3,cmdS4 . I tested for error producing situation but no rollbak occured. How can I solve this problem. Thanks. Below is my SP . .......... .......... .......... begin transaction .......... ..........If Len(ltrim(rtrim(@cmdS)))>0 execute(@cmdS) If Len(ltrim(rtrim(@cmdS2)))>0 execute(@cmdS2) If Len(ltrim(rtrim(@cmdS3)))>0 execute(@cmdS3) If Len(ltrim(rtrim(@cmdS4)))>0 execute(@cmdS4) If Len(ltrim(rtrim(@cmdS)))>0 or Len(ltrim(rtrim(@cmdS2)))>0 or Len(ltrim(rtrim(@cmdS3)))>0 or Len(ltrim(rtrim(@cmdS4)))>0 Beginif @@ERROR <>0 begin rollback transaction set @Hata = 'Error !' end Else Beginset @Hata = 'Sucessfully executed :)' End End commit transaction RETURN
I just made a transaction which writes and reads some data from a database as one unit. Now, I would like to know, what happens, if this transaction does not commit, but instead performs a rollback? Should I use a boolean loop which keeps on trying the transaction until it commits or is there an alternative? I would appreciate any suggestions or help!
I am just starting to falmiliarize myself with SQL transactions... I just created an SQL transaction... The first statement gets a value, if the value equals "" then the second statement executes... So if the value <> "" Then the second statement wont execute... What would happen in this scenario if the .Rollback is triggered? Heres my code:
Try conSqlConnect.Open() objTransaction = conSqlConnect.BeginTransaction cmdSelect.Transaction = objTransaction cmdInsert.Transaction = objTransaction dtrdatareader = cmdSelect.ExecuteReader() While dtrdatareader.Read() varCheckNumber1 = dtrdatareader("Status") End While dtrdatareader.Close() If varCheckNumber1 = "" Then cmdInsert.ExecuteNonQuery() End If objTransaction.Commit Catch objTransaction.RollBack Return "00" Finally If conSQLConnect.State = ConnectionState.Open Then conSqlConnect.Close() End If End Try
I've some problem in SQL Server 7.0 and hope some one out there will help me with it. The problem is as follows:
When I view a proces in the process info screen, I see one process says "Rollback" in the status window and the spid is greyed out like a system process. In the command window it says, Select and wait type is "Exchange". This is taking too much cpu cycles and causing much of the server to perform very slow.
I tried to kill this session from the ISQL and Enterprise manager, But to still see this process running on with out getting killed. Also this process keeps coming.
Please advice me how to get rid of this and why is it ocurring again and again.
I am getting an automatic rollback in a transaction that fails to insert a row, via a trigger, due to a duplicate key. The transaction is automatically rolled back and unless the error correctly handled a 'partial update' occurs. Why in the example below is the transaction automatically rolled back (statements 1 to 3) and is there any way of knowing what errors generate an auto rollback. For example if the duplicate key error occurs on the insert into table B below, and not the triggered update, then no auto rollback occurs.
Begin Tran -- 1 insert into A values ('A1') -- 2 updates OK insert into B values ('B1') -- 3 updates OK -- triggers insert into B_Hist -- 4 Fails with Duplicate Key error insert into C values ('C1') -- 5 updates OK Commit Tran -- 6 3902 error No corresponding Begin
I have this weird situation on one of the production database. Somebody ran a sql script couple days ago and must have hit cancelled. Somehow this script is still rolling back. The script is actually very simple statement..should not take more than couple miliseconds to run. Is there any way to see whether rollback is really in process or this process has simply hung? is there any other way to get rid of this rollback without stopping and restarting sql server services? Can it cause problems in getting sql server up again if I restart sql server services?
This may sound a little silly, but does anyone have any words of wisdom on how to optimize a server/database for minimim rollback? We have some multimillion row tables we were trying to do updates against, and after several days they increased the size of the transaction log to the point they filled up the drive the database files/logs were on. We've now been running a rollback for about five days. I'd like to make sure this doesn't happen again.
In MOC (SQL 7.0 Implemetation), it says there will be rollback, but in my humbel opinion, there should be nothing happened.
I think there was no dirty writing in this case. No data pages written, no log pages written to disk. Am I right? Then, there might be nothing happened instead of ROLLBACK.
If I am wrong, would you please let me know what is wrong in my thougt?
If you have a begin tran in a sp that calls another sp and that called sp fails. can you execute a roll back from the called sp, or should it return and error which will roll back the begin tran in the original sp
I have been informed that most (if not some) RDBMS are faced with problems with the SQL ROLLBACK (also known as undo) command. I have been searching online, but have not found any useful relevant sources on what this issues are.
I am wondering if their are any experienced MS SQL Server (or any other RDBMS) users who might be able to provide feedback on this matter.
Don't ask, but I iadvertantly managed to delete 57,000+ records in the EM console and I need to roll back the database a few minutes to recover said information. I do not care about transactions that have occured since due to shutting the system down immediately upon realizing what I did. I have however never had the need to roll back the database before, so I have experience doing so. Any help is desperately needed and greatly appreciated!!!
I am not sure about the ANSI/ISO rules on this? We have a SEQUENCE and insert it into a table.
CREATE SEQUENCE Cheque_Seq AS INTEGER START WITH 1 INCREMENT BY 1; CREATE TABLE Cheque_Book (cheque_seq INTEGER*NOT NULL PRIMARY KEY, cheque_amt DECIMAL (15,2) DEFAULT 0.00 NOT NULL);
If I do an insertion with the SEQUENCE in a transaction and then ROLLBACK the transaction, what happens to the sequence value? My thought is that the value should reset to the state it was in before the transaction (basic definition of a ROLLBACK). But it does not in the products I have looked at.
I tried adding columns to a table on a database manually. Unfortunately the database doesnt have a lot of free space, and on saving the columns, it was slowly running out of space. I made the decision to end the (sql enterprise) task via task manager. I am now unable to get into enterprise... I am assuming it is trying to rollback?<>?. can anyone tell me what happens next. will the database still be intact..
