TheSQLGuru (1/26/2011)

---

shaun.stuart (1/26/2011)

---

You generally don't want any database files on the same drive as the OS because database files can grow. If something goes wacky, they could grow to consume...

Thanks Lynn. However can you explain why? If the performance of the spindles are fine, why use up disks that can help the database performance by adding to a RAID10...

Fyi - we're tracking stats on the servers. below are the ones for server6 and below that are server9.

SERVER6:

Average Latch Wait Time (ms)51.227

Average Wait Time...

ALZDBA (1/20/2011)

---

shifty1981 (1/19/2011)

---

...

I have been told that rebuilding our indexes typically takes about 2 hours and that it puts a pretty heft load on the server. However I hoping to...

Hi everyone. sorry for the delayed responses. Was gone for a while and then had to catch up at work. Much to give an update on. I'll try and do...

sqlbuddy123 (1/7/2011)

---

I think the problem could be with the index fragmentation and missing indexes and also may be the disk fragmentation too.

Disk Metrics for K drive look pretty bad. Full...

Craig Farrell (1/7/2011)

---

Jason,

%DiskTime and Queues are no longer valid parameters if you're working against SAN systems. If you have independent RAIDs, that's different. They're really meant for single...

jswong05 (1/7/2011)

---

Hi, shifty1981,

Did you check FreeListStalls ? FreePages? MemoryGrantPending ?

Did you have a lot read-ahead reads?

Don't knwo what these are? Are they in perfmon? Is there a good article...

Craig Farrell (1/7/2011)

---

Shifty,

Do get the wait stat information, it's important, but this also concerned me:

Average disk sec/Trans

P drive (RAID10 - 14 drives) 0.014

G drive (RAID10 - 14 drives) 0.014

K drive...

Here are the results:

the K drive is on a SAN. yeah I'm not too excited about those stats. it's weird though, we have an individual san just like that for...

GSquared (1/7/2011)

---

Page Life Expectancy is the number of seconds data is left in RAM before SQL Server needs to move it out to make room for more data.

The Buffer Hit...

Ninja's_RGR'us (1/7/2011)

---

shifty1981 (1/7/2011)

---

Ninja's_RGR'us (1/7/2011)

---

Well the GB left is the GB left. As you know, sql will grab all it can.

So if you have plenty of ram left after the...

Steve Jones - SSC Editor (1/7/2011)

---

Your Target server memory is what SQL wants, and it matches your total. So I don't think you have memory pressure here.

The high I/Os might...

Ninja's_RGR'us (1/7/2011)

---

Well the GB left is the GB left. As you know, sql will grab all it can.

So if you have plenty of ram left after the server's been...

Steve Jones - SSC Editor (1/7/2011)

---

I might check on a few things here.

1. Check the buffer cache hit ratio, and see if it is high. Below 90%, I'd definitely...