Recent content by tombos

I'm in the process of looking hard at the execution plan of a stored procedure to optimize access to a 13M record table, and I'm seeing something odd. When looking at the cost of each subtree in the proc, the results for the following statement have stumped me: select col1, col2, etc from...

We've built our first ASP.NET application on a Windows 2003 box running all the latest service packs. It hits an SQL 7 server for data. Both boxes are currently inside the firewall in an effort to eliminate the firewall from the equation. We periodically receive an Internal Server Error...

Can anyone recommend a good on-line or CD-based training course for ASP.NET that I can purchase (or even better if free) and work through at my own pace? I've made a huge amount of progress on my own, but I'm seeing that some things--especially related to database access--have several different...

I have a family of fat client apps in our company that were developed in VB6 and hit SQL server for all the database access. For a variety of reasons, I want our next generation of applications to be browser based so that it's easier to run them across the WAN, Internet, etc. So... I suppose...

In SQL 7 with the latest service packs, I have a stored procedure (actually several) that uses the following type of statement: UPDATE Table1 SET a.col_abc=b.col_def, a.col_stu=b.col_xyz FROM Table1 a, Table2 b WHERE a.col_pqr=b.col_pqr Of course, all those column names are arbitrary. Bottom...

You're probably not going to get many takers on this. I've never done what you're attempting, but I have replaced one system that was precisely what you're describing. It was incredibly unreliable and slow. Of course, that could have been the implementation, but there appeared to be some real...

By any chance, did you change the name of the server AFTER you installed SQL? I did that once and had to re-install SQL 7 to make the service start. I can tell you that I run SQL 7 under Win2K on two different servers, and it's been extremely reliable.

Well, I have a new twist today that will sound crazy. I promise that I'm generally competent. After I read your previous posting, I went and did the Update Statistics and the DBCC DBREINDEX commands. (I would also note that the indexes are rebuilt every Sunday as part of our maintenance plan.)...

Yeah, I've updated the statistics and reindexed too. Nothing changed in the behavior. I'm simply confused why SQL would decide the query is best executed with a table scan for one value but that using the index is best for the other. Is there any way that I can FORCE it to use the index when...

I have a table with approximately 800K rows against which I run a common query. For example: select zip,route_number,city,hh_count from tblRoutes where event_id='MS020730' and market_code='AL-ALB' and job_no='001' What's causing me trouble is that SQL chooses to execute this query in...

We've experienced a situation on more than one occasion where a given process on our SQL Server ends up locking 1000-1200 different records and/or pages in a database. The last T-SQL statement that executed was SELECT * FROM vw_ViewName. That statement was issued from a VB6 application using...

The easiest way I know to do this is to set a trigger that fires on INSERT, UPDATE and DELETE operations for that table. Within the trigger, you could identify the changed columns in the table and insert records into a separate log file.

I'm using a trigger to keep a quickly accessible sum updated on a master table so that I don't have to query the child table every time I display the total. I want to do this within a trigger on the child table. However, whenever the trigger fires for a multiple-row transaction, i.e. the...

You could also use something like dateadd(d,1,getdate()), which would retrieve tomorrow. Likewise, dateadd(d,-1,getdate()) would retrieve yesterday.