stop the merry-go-round

There is an acronym lookup page, and not only does it tell you the acronym and all of the overloads for it, but it's broken down by business unit so you can map acronyms to your unit's nomenclature

Josh Darnell

12:37 PM

@PaulWhite I don't see anything about pivot / unpivot in there :^)

I did have a hard time getting through that post.

It seems very practical.

Paul White

I ruined the transcript ha ha

Josh Darnell

@SeanGallardy LOL that's fantastic. I worked for Duke / Progress Energy for a while, and they had that exact same thing.

@PaulWhite :o

Paul White

@JoshDarnell You have mail

Josh Darnell

@PaulWhite I see!

Paul White

Time travel by email

Josh Darnell

12:42 PM

Fancy.

Which suggestion did you provide for this series?

Paul White

Buggered if I can remember

Wasn't pivot or unpivot that's for sure

By the way I sent that email using the Advanced Delivery System

Josh Darnell

I thought something seemed different about it.

I was also warned it might be a phishing attempt.

Paul White

Really?

How rude

Also, please send me your fish

Josh Darnell

> BE CAREFUL. Paul White has never emailed you from this address

Paul White

Sage advice

Also, I need a herb recommendation to go with the fish

Josh Darnell

12:47 PM

I don't have thyme for that right now.

Paul White

Forrest is that you

Peter Vandivier

what is the dill with this chat

Paul White

It's tarragon downhill for sure

Josh Darnell

If only someone could keep these puns at bay.

Paul White

lol

Peter Vandivier

12:49 PM

it used to be anise chat

Paul White

I just had to star anise

Josh Darnell

Paul just dunked on Peter's pun.

Paul White

51 minutes into Friday and I've peaked already

Josh Darnell

It was a good run.

Peter Vandivier

i'll try not to get salty about it

Sean Gallardy

1:06 PM

Wow.

I leave you all unsupervised for 15 minutes...

Erik Darling

i have nothing oregano to add

@Forrest email

Josh Darnell

1:28 PM

Sounds like an email problem.

Joe Obbish

1:57 PM

does Paul eat fish?

1 hour later…

George.Palacios

3:22 PM

smoked paprika

did I do it right?

Erik Darling

3:39 PM

@JoeObbish fish and eggs!

@George.Palacios mom's spaghetti

Forrest

4:02 PM

I just finished my couple-hours-notice presentation to devs. In part of it, I showed a query that ran in 0ms, and a bad form of that query that ran in 1:50.

I asked them, "based on math, how many times slower is the bad query?"

The answer I got: "150 times"

Erik Darling

4:17 PM

Why is that wrong?

Forrest

4:36 PM

🤨

Erik Darling

Those magnificent geniuses removed the colon and gave you a number. That's the most efficient answer.

Joe Obbish

the best developers are lazy

150 times is close enough to the correct answer anyway

Paul White

5:29 PM

The maths community will be pleased to hear that division by zero has been solved anyway

Erik Darling

6:16 PM

@PaulWhite bro just use nullif

@JoeObbish wisdom

Paul White

6:46 PM

@ErikDarling oh thanks I always wondered what null was for

1 hour later…

Joe Obbish

8:05 PM

looking at a database with RCSI

clustered index scan with TABLOCKX takes 25 seconds, without the hint takes 120 seconds

no blocking in the wait stats

I suspect the version store is playing a role, or at the very least I want to rule it out

I can't figure out how to do that though

not sure what else it could be

it's an ordered parallel scan

Josh Darnell

Do you version store DMVs provide any clues?

This reminds me of the Q&A from that other dude from Madison.

A: Long Running Query on Read-Only Replica that takes moments on the Primary

This answer is in addition to Joe's answer as I can't be 100% certain it is the version store, however there is enough evidence so far to imply that to be part of the issue. When a secondary replica is marked as readable a good steady state for versioning information needs to first be attained s...

Joe Obbish

@JoshDarnell don't know what to look for

ETW is a bust because the server is at 50% CPU without me running a test query

Josh Darnell

Is the difference in the longer running one all CPU?

> The walking of the chain is almost a purely cpu operation, so when the chains become very long and the amount of rows returned is very high, you get a (not linear, but can be close) increase in time for the query.

A clearer question would be: For the 120 second query, is CPU time close to 95 seconds * DOP higher than the 25 second query?

Joe Obbish

9:12 PM

@JoshDarnell are you asking if the queries have the same CPU time to elapsed time ratio?

Sean Gallardy

9:25 PM

No one mentioned RC1 at all

Joe Obbish

@SeanGallardy WRONG, read the transcript

you can also look at starred messages

Sean Gallardy

@JoeObbish you should know me better by now

Paul White

1 hour ago, by Joe Obbish

looking at a database with RCSI

HtH

1 hour later…

Joe Obbish

11:02 PM

@PaulWhite I don't understand

almost got hit by a bus a little bit ago

that would have solved a lot of problems

Paul White

For Sean’s benefit

I thought he typo’d RCSI

Joe Obbish

"drivers must: Yield to pedestrians who are crossing the highway within a marked or unmarked crosswalk at an intersection where there are no traffic lights or control signals "

seems pretty cut and dry