Evil Kraig F (9/8/2011)

---

When is the process being run? Which one has more memory available? Are you transforming in the package or after drop to staging tables? ...

PaulB-TheOneAndOnly (8/23/2011)

---

Shurkadze (8/22/2011)

---

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/22/2011)

---

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/22/2011)

---

The DW contains small subset of columns of OLTP. Relying on timestamp may not be meaningful since only a small subset of columns changed...

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/22/2011)

---

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/22/2011)

---

The DW contains small subset of columns of OLTP. Relying on timestamp may not be meaningful since only a small subset of columns changed should trigger...

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/22/2011)

---

The DW contains small subset of columns of OLTP. Relying on timestamp may not be meaningful since only a small subset of columns changed should trigger load into...

Lowell (8/22/2011)

---

Shurkadze (8/22/2011)

---

What if you have a row modified in OLTP with a column value changed that is not mapped to DW. The data then should not be moved to...

PaulB-TheOneAndOnly (8/22/2011)

---

Shurkadze (8/19/2011)

---

We have a transactional database with every table containing create_date and update_date columns.

We designed a data warehouse and developed ETL process (initial and incremental).

Is it a valid concern...

LutzM (7/8/2011)

---

To repeat what I stated earlier:

It seems like the indexes available can't be used in the first statement since the indexes are not covering indexes (e.g. an index for...

roshan.zanwar (7/7/2011)

---

Obviously, COUNT(*) will return u the result faster as it just counts number of rows.

When using count(*) you are just joining the tables & counting number of rows.

While in...

How fast does it return results when you replace the update with a select ?

Are there Nulls in the columns on which your joining the tables.

When I run COUNT (*)...

Does the file need to be imported as 1 column in 1 row and then parced out or

is there a way to import the file parsed out into separate columns/...

You are an amazing person and an inspiration to all of us.

Wish you speady recovery and keep your great spirit UP!

Alexander.