GilaMonster (9/27/2011)

---

It probably...

From my experience with this client, and admittedly it is my first major dealing with heaps, I see that a high percentage affects performance adversely. As to how high...

GilaMonster (9/27/2011)

---

opc.three (9/27/2011)

---

mister.magoo (9/27/2011)

---

opc.three (9/27/2011)

---

Did you use an INT for your identity column? Try it with BIGINT. I have a theory that since a RID is 8 bytes that...

mister.magoo (9/27/2011)

---

opc.three (9/27/2011)

---

Did you use an INT for your identity column? Try it with BIGINT. I have a theory that since a RID is 8 bytes that using a...

GilaMonster (9/27/2011)

---

opc.three (9/27/2011)

---

Disabled indexes are not maintained...unless BOL is wrong which is something Gail has compelled me to consider given all the times she has cited incorrect docs 😉 That...

mister.magoo (9/27/2011)

---

opc.three (9/26/2011)

---

@mm Something occurred to me, you may still have heavy fragmentation from the continuous addition of new rows and the deleting of existing rows over time. Here is...

Sorry Grant and Gail if I am being dense, but I am still not sure I got an answer for my original question. Maybe I am incorrect, but it seems...

@mm Something occurred to me, you may still have heavy fragmentation from the continuous addition of new rows and the deleting of existing rows over time. Here is what I...

No. If you are wondering whether it suffers from the...

Edit: blah

mister.magoo (9/25/2011)

---

Thanks for the info.

My understanding is that you only get forwarding entries if an update changes the size of the row such that it won't fit in the existing...

mister.magoo (9/25/2011)

---

Hi,

I also have to work with a third part OLTP that uses all heaps (It sounds just like yours - historically coming from C-ISAM), and was just wondering what...

PS maybe "double work" is not the right term...how about "extra work" updating presumably fragmented non-clustered indexes

GilaMonster (9/25/2011)

---

opc.three (9/23/2011)

---

I am wondering if anyone knows if during a heap rebuild if the non-clustered indexes are updated with the new RIDs as the heap is being defragmented...

Thanks for the response Grant.

Heaps are generally weak tables. They have a place in design, but...