The plan does show the cardinality of the table to be the same.

It's the actual rows coming out of the scan that are so wildly different.

From a quick glance it...

I thought it was stated that client_id in the client table was varchar? In that case the LIKE '%.%' isn't an issue.

The bit pointed out by Chris is the main...

There are few considerations here, and I think some of them are running together in the conversation.

1) Assuming the same CPU architecture, sure, 4 logical cores backed by 4 physical...

Lynn Pettis (11/17/2016)

---

Jacob Wilkins (11/17/2016)

---

Lynn Pettis (11/17/2016)

---

Jacob Wilkins (11/17/2016)

---

Also, print out the value of @sql instead of trying to execute it.

You'll see there are several syntax errors.

One of the main...

Lynn Pettis (11/17/2016)

---

Jacob Wilkins (11/17/2016)

---

Also, print out the value of @sql instead of trying to execute it.

You'll see there are several syntax errors.

One of the main ones is that you're...

Also, print out the value of @sql instead of trying to execute it.

You'll see there are several syntax errors.

One of the main ones is that you're trying to use COUNT(*)...

It means somebody wanted to make sure that the createid stayed the original and that the updateid was the last person to update it.

That's fairly reasonable, and there could be...

Most likely an update trigger on the table. I'd check that first.

Cheers!

Well, you could...

Here's another method:

SELECT CaseID,

min_date=COALESCE(CASE WHEN MTDThresholdDate<YTDThresholdDate

...

For mirroring there's a perfmon counter, "Log Bytes Sent/sec" that exists for each database. Despite the "/sec", it's a cumulative counter (Perfmon just automatically does the calculations for you).

You...

Is SCCOE_Order_Progression the database you're running all this in? Wondering if there's a trigger on that table inserting into a table of the same name in another DB.

Far-fetched, but so...

CREATE TABLE [dbo].[data_load]

(

[ID] [INT] NOT NULL,

[PC] [INT] NOT NULL,

[AM_System_Type] [VARCHAR](50) NOT NULL,

[PeriodKey] [VARCHAR](50) NOT NULL,

[File_Path] [VARCHAR](4000) NOT NULL,

[Fale_Name] [VARCHAR](255) NOT NULL,

[File_Type] [VARCHAR](50) NOT NULL,

[Status] [VARCHAR](255) NULL,

[PlaceHolder1] [VARCHAR](4000) NULL,

[PlaceHolder2] [VARCHAR](4000) NULL,

[SysStartTime]...

For a little bit about what determines if a plan stays in cache, you could take a look at https://technet.microsoft.com/en-us/library/ms181055(v=sql.105).aspx#Anchor_0.

I definitely wouldn't just assume that you're neither pushing plans out...

Here's another 2008-friendly way that only accesses the table once. That comes at the cost of a couple extra sorts, and whether the extra IO or the extra sorting is...