stop the merry-go-round

is anyone able to parse this sentence? "Using this trace flag, it is not recommended to execute queries specifying a DOP over the number of schedulers in a single, with queries specifying a DOP below or equal to the number of schedulers in a single node."

jadarnel27

*single node

Joe Obbish

but even then it doesn't make sense?

jadarnel27

Oh that.

Yeah that's bizarre.

Joe Obbish

"Using this trace flag, it is not recommended to execute queries specifying a DOP over the number of schedulers in a single node, [as this could interfere with] queries specifying a DOP below or equal to the number of schedulers in a single node."

maybe???

jadarnel27

Yeah, I think it's saying you shouldn't run the high DOP query simultaneously with the lower DOP query.

Joe Obbish

3:09 AM

lol

I run a MAXDOP 24 query with 19 other MAXDOP 8 queries at the same time 😎 😎 😎 😎 😎

am I too cool for school?

jadarnel27

It sounds like, at the least, you are too cool for TF 2467.

Joe Obbish

no, that's with TF 2467

jadarnel27

Well that stop that! You are failing to comply with the documented recommendations from Microsoft.

Kids these days.

Joe Obbish

yeah it's pretty funny

it's finally documented and we immediately aren't complying with the documentation

jadarnel27

Pedro added that to the docs 3 weeks ago. You should send him a thank you note.

Joe Obbish

3:15 AM

wow, that long ago

I already did as I was the one who asked him to document it

jadarnel27

Well, it was committed in GitHub 3 weeks ago. It takes a while to show up on the website.

Joe Obbish

I'd try to fix the typo but I don't know what the intended meeting is and I also don't want to learn how to use github

just ask Erik

9 hours later…

sp_BlitzErik

12:18 PM

jeez, late night for you

2 hours later…

sp_BlitzErik

2:14 PM

Q: TSQL compare the values of the same column cells without knowing their values

I have a table which contains just two columns, it is X and Y type of int. Column | Type X | int Y | int This are coordinates. I should write a procedure now while I have no access to the information contained in the columns now, so I can't select x where y = 'somevalue' cause I do...

t-sql

COLLUSION
O
L
L
U
S
I
O
N

Joe Obbish

2:33 PM

@sp_BlitzErik NO COLLUSION

Paul White

3:03 PM

follow up question:

-1

Q: solving problem of row comparison in tsql

I am trying to solve this problem This is my code create table #tbl ( x int, y int, rn int -- row number ) -- order by X ASC before inserting into #tbl insert into #tbl select x, y, row_number() over (order by x) from functions order by x asc -- select * from #tbl declare @maxrn int...

sql-server t-sql

might want to put on 😎

Joe Obbish

the more questions the merrier

that's what they say, I think

seems like async stats burned me real bad

Paul White

oh ... what was the bigint bitmap problem in the end?

Joe Obbish

@PaulWhite the query referenced a view in the dbo schema. so the query with the default schema ended up looking at the dbo versions instead of at the special bigint schema

so we were joining a NUMERIC(18,0) to a BIGINT column

(to do this test I created copies of all tables with BIGINT instead of NUMERIC(18,0) and threw them in a different schema)

Paul White

oh I see!

I hadn't realized those two things were connected

sneaky

always schema-qualify!

Joe Obbish

this app has problems with that

with that said, it's convenient in this case not to schema qualify

we can use the same test queries instead of having to modify them all

another test has thousand of queries

I don't know how difficult it would be to parse all of those queries to change the schema (if we had specified dbo)

Paul White

3:19 PM

use synonyms?

idk perhaps you have the best solution for your needs already but ouch

Joe Obbish

"ouch" to which part?

and no, I doubt it's a good solution

Paul White

ouch is my reaction when bitten by things like that

an empathetic ouch if you will

Joe Obbish

now that I think about it, there's a part of the application that takes advantage of the lack of specified schemas

good times

Paul White

everything's a compromise I guess

Joe Obbish

the whole thing is silly anyway

we should have picked BIGINT 15 years ago

then I wouldn't be struggling through this

the code to recreate all of the tables and indexes took about 3 days off and on

then I had to recreate all of the views

now I think it's mostly good

Paul White

3:23 PM

as you say, good times

I know all about compromise today, having finally given in and installed tampermonkey today so I can run userscripts on SE sites.

I swore I would never do it

Joe Obbish

why were you against it?

Paul White

mostly because (a) I don't know how to write the code, and therefore have to put a lot of trust in the author; and (b) once userscripts are active, one can never be sure if a bug is caused by SE or the script

I dislike layers of complexity in general, unless I understand them really well

Joe Obbish

reasonable concerns

are you using the "delete all the comments" script?

Paul White

ha ha is there one of those

no, but I cannot live with links always being underscored

so to avoid hating the sites, I have a script to remove the underlines

Joe Obbish

it seems like you do a lot of work on other sites. what kind of work is it? don't they have their own moderators?

Paul White

3:34 PM

I read the main metas a lot to keep up with developments

the new theme will be coming to Database Administrators in due course

there are other reasons as well, since moderators talk to each other about stuff across the network

sp_BlitzErik

4:01 PM

nice chat you're having

jadarnel27

I don't mind the underlines really.

But I've always been very passive about the UI of websites / apps that I use.

sp_BlitzErik

4:26 PM

web dev doesn't care about design, story at 11

jadarnel27

Haha well.

Good point I guess.

I care about design when I'm the one designing.

sp_BlitzErik

v. selfish

be the design you want to see in the world

jadarnel27

That's deep.

Joe Obbish

4:50 PM

@sp_BlitzErik brutal as

1 hour later…

jadarnel27

6:07 PM

I tried to create a scenario last night where tons of competing locks in a database were taken causing the lock monitor to use a lot of CPU. I was very unsuccessful.

Turned off lock escalation on the Posts table and was able to get tons of individual locks taken.

The deadlock monitor, though, was not impressed.

Joe Obbish

have seen tons of locks put CPU pressure on transaction log writers

jadarnel27

Interesting.

bbl, haircut time.

sp_BlitzErik

6:59 PM

@jadarnel27 don't make the same mistake i did

Joe Obbish

7:21 PM

you people are so vain

spending all of this time and money on haircuts

sp_BlitzErik

7:35 PM

some of us need to be presentable

we can't all walk around looking like a struggle against humanity

Joe Obbish

8:27 PM

the married guy who works from home claims a need to look presentable

jadarnel27

@sp_BlitzErik that's a killshot

@JoeObbish do you happen to remember anything about that workload? Like how I could repro that problem?

sp_BlitzErik

@JoeObbish i'm on webex 3-5 days a week. how i look from the ribs up is v. important.

Joe Obbish

@jadarnel27 yeah, get a real server

@sp_BlitzErik just use a recording

one of these days you're going to fix sql server issues while wearing a suit

sp_BlitzErik

8:45 PM

are you gonna buy me one?

Joe Obbish

we all know you already own one

you live in NYC

sp_BlitzErik

not i

i had one from a wedding five years ago that we sold

Joe Obbish

I bet they stop and frisk you daily

jadarnel27

@JoeObbish extremely unwelcoming!

jadarnel27

9:11 PM

But for real, if you have any suggestions, that would be neat. Will take another stab at it this evening.

Joe Obbish

9:27 PM

@jadarnel27 bruh

you're the one who asked if I knew anything about my workload

I know more about my workload than you know about your children

jadarnel27

9:38 PM

Um.

Okay?

sp_BlitzErik

9:57 PM

all joe knows it that it's virtualized

1 hour later…

Joe Obbish

11:11 PM

@sp_BlitzErik you're virtualized

sp_BlitzErik

@JoeObbish your waifu is virtualized

sp_BlitzErik

11:39 PM

Issue with security update for the Remote Code Execution vulnerability in SQL Server 2016 SP2 (CU): August 14, 2018 blogs.msdn.microsoft.com/sqlreleaseservices/…

> For one of those releases, SQL Server 16 SP2 CU (KB4293807), we inadvertently published additional undocumented trace flags that are normally not on by default.

Joe Obbish

I wonder what they do

« first day (199 days earlier) ← previous day next day → last day (490 days later) »

all rooms