« first day (4994 days earlier)      last day (97 days later) » 

00:29
@ErikDarling What happened to PlanA?
@ErikDarling That might be the least informative video ever made. Why even present that in video form
01:08
I didn’t watch it. I was on an airplane.
There was a slim chance that it would be the most exciting SQL Server enhancement since uh
Oh give me a minute
That thing with the maps
 
2 hours later…
02:52
Oh you mean Service Broker
 
2 hours later…
05:11
Wordle 1,175 4/6*

⬛🟨⬛⬛⬛
🟩🟩⬛⬛⬛
🟩🟩🟨⬛⬛
🟩🟩🟩🟩🟩
 
4 hours later…
09:07
Wordle 1,175 5/6*

⬛⬛⬛⬛🟨
⬛🟨🟨⬛⬛
🟩🟩⬛🟩⬛
🟩🟩⬛🟩🟨
🟩🟩🟩🟩🟩
@ErikDarling too localized
@JoshDarnell I read the article and think the title should be changed to: "Add confusion....".
09:48
Wordle 1,175 4/6*

⬛🟨⬛⬛⬛
⬛🟨🟨⬛⬛
🟨⬛⬛🟨🟨
🟩🟩🟩🟩🟩
 
4 hours later…
13:46
Wordle 1,175 5/6

⬜🟨⬜⬜🟨
⬜⬜⬜🟨🟨
⬜🟩⬜🟨⬜
🟨🟩🟩⬜⬜
🟩🟩🟩🟩🟩
14:39
Azure is expensive and slow, and the fine-grained access control puts gatekeepers all over the place. Unless your devops function is really on the ball, you're forever waiting a week to get 5 minute configuration jobs done.

So, therefore, I have to conclude that it doesn't deliver on price, performance or flexibility.
@ConcernedOfTunbridgeWells You're not wrong.
I think a big wintel server with an EPYC chip and a few hundred GB of RAM would get you a private cloud for about the same amount as it would cost to rent the same capacity on Azure for a month.
And it would be faster and more flexible.
Is AWS any cheaper?
Not really, it's generally slightly more expensive (AWS)
@ConcernedOfTunbridgeWells This is exactly correct. There are some CIO posts about them doing this for their company and saving tens of millions per year.
I get it if you have no staff or expertise, but these places using cloud in the tens of millions per month are throwing money away, it's crazy.
David Hannemeier Hanson was rabbiting about this on t'interwebs a few months ago. Apparently they bought about 500k worth of server kit and moved all their hosted basecamp instances onto it, saving about $1.5m/year on hosting costs.
Also, from the items I see, the same companies think they don't need a DBA (for example for databases [SQL/PG/MySQL/etc]) and then make support tickets because they aren't getting good perf from their crap queries and bad schema designs expecting the hosting company to be their DBA... and then when they get told they aren't their DBA, I can only imagine the look on their faces.
14:47
@SeanGallardy I also think the sheer volume of stuff to twiddle on Azure makes devops a significant cost in running an Azure estate anyway. Plenty of rumblings about with customers being unhappy about the lack of cost savings.
My favorite are the ones where the region/DC has an issue and the SLA is 99.9%, they are down for a day or two and complain. You signed a 99.9%, that's well within it. If you don't want downtime, ruin your own.
@ConcernedOfTunbridgeWells Correct, but let's be honest, you don't go to the cloud for cost savings. It's abhorrently more expensive to run cloud services for 1 year than to buy the requisite infrastructure.
My view is, if you want 5 nines, get a mainframe. If you're serious about that SLA the cost of the testing and change control far outweigh the hardware costs.
Penny wise and a pound foolish
@SeanGallardy My view is - if not cost then why. It's an objectively worse platform than on-prem Windows or Linux in just about every measure.
Painfully slow I/O. Very expensive CPU. Overly complex security and deployment because it has to be all things to all people.
Because it's the rage in the CIO world - we're in "the cloud" like the board wanted. Same thing is now with "AI". Remember when all the rage was "ML" and "BlockChain"? It isn't pushed because they need it, it's done because it sounds good.
Great Silicon Valley episode that makes fun of this very thing.
14:50
It's definitely flavour of the month. Look back and we have Web 2.0, semantic web, big data, ~data mining~ predictive analytics . . .
@ConcernedOfTunbridgeWells I've been told to get ready for Web 3.0
Web 2.0 was a bunch of apps on a device that just take you to the same webpage as a browser.
Every technology has its killer apps. Blockchain, for example, has found its niche in crypto speculation and buying weed.
Agreed, I'm all about the right tool for the right job.
Generative AI will find its niche in automated CV writing and pornography,
Big data found its niche. Actually, I can't think of much if anything useful that big data did except sell advertising.
Predictive analytics does market segmentation - in exactly the way that data mining used to do 35 years ago. Mainly because it's just the same algorithms with a bit of lip gloss added.
I do believe the cloud has a place, especially in small to medium business environments. For example, cloud hosted Email is pretty decent in price and performance.
14:53
There are certainly uses for these things. SAAS in one form or another is certainly usefuk.
Exactly. Right tool for the right job.
Workday, Salesforce and their ilk do a pretty good job of what they do.
OTOH, I've seen several lakehouse projects that read the same structured data and produced the same financial MI as their predecessors. And they all had performance issues because spark dataframes can't do efficient joins.
This could be fixed by a distributed hash lookup or index capability that was suitably efficient.
I suppose things have changed in the past 20 years. In the just under 20 years I've been in Blighty I've grown into a grumpy old man.
@SeanGallardy I love Silicon Valley. It's due some rewatch
15:05
Haha I was thinking the same thing, it was such a good series. I also have some interesting memories tied to it as I was literally in silicon valley when it first aired and started watching it.
15:46
were you disrupting the space?
 
1 hour later…
17:00
I'm hoping @SeanGallardy can clear this one for me
0
Q: Database both joined and not joined to an AG

ZikatoI have a Distributed Availability Group (DAG) from cluster Main to cluster Alt - both with two replicas. The Main is the primary and Alt is a forwarder One (of many) databases called MyDb in the AG on the forwarder side's secondary replica was in a bad state and was dropped. I've used automatic s...

17:37
@ErikDarling is on to something. The master database is set to compatibility 150 (SQL 2019) which makes sense since this is a SQL Server 2019 instance. The CRM database is set to 140 (SQL 2017). — Ken 3 hours ago
HE SHOOTS HE SCORES
18:36
OP adds plans in comments
Every day is an uphill struggle
@Zikato Can you put the output of the dm_hadr_database_replica_cluster_states for that database? Hide any PII.
> Microsoft SQL Server 2022 (RTM) - 16.0.1000.6 (X64)
MFW
18:53
@SeanGallardy cannot anymore. A teammate removed the DB from the forwarder and the secondary and that fixed it. I'm still interested in what's causing this though.
If there's no data to collect then there's no way to really RCA it unless the errorlog has enough in it... which is rarely the case for HA internal related items.
Unless you have a repro on NON-RTM
Ex: something on the latest 1-2 CUs
I don't have any lab environments for AGs let alone DAGs
You can always Email me if it's not good for chat
We seem to be avoiding the RTM issue
And by 'we', I mean @Z
I don't think it's relevant, I've seen this on other versions
19:01
@PaulWhite Ooooooooooooh.
Sick burn.
does RTM mean "read the manual"?
The joke is RTM could mean Release To Manufacturer OR Read The Manual
You probably don't think manuals are relevant
I'm still learning to read
Docs Learn won't help you
Chat markup correct on the first try! Go me!
19:05
🥳
Might be a day to play the odds at the lottery
I think you used all your luck
Anyway, I hear AGs are 99.9% available
But how are the kids going to go in there, it's too small
@PaulWhite You're probably right
@SeanGallardy it will be 3 times bigger than this
19:09
That movie was hilarious
A comment with ten upvotes saying the OP should add attribution for the quoted answer portion that helped them. Now a moderator flag. I spent 0.00007s editing the attribution into the OP's answer. People, idk.
Maybe the obvious solution is too obvious sometimes
@PaulWhite practice makes perfect
On a six-year-old answer as well
It's a good job I attended Erik's Permanently Optimistic class
aka POC
I got a gold star
and a certificate
19:18
I wonder if that's his stance on concurrency levels, too
He does like RCSI
@PaulWhite how much alcohol was involved?
I can't really parse the question
Can you be asking if we left any spare?
All of it, then
Like it needs saying
Anyone up for a game of skribbl?
Is that eastern bloc scrabble?
19:28
No. One person gets a choice of 3 words to draw and others are guessing. Then it switches around round-robin
Nice, I didn't know it was called skribbl. Here it's called pictionary.
Well, skribble is the online version. I'll keep this room open for 30 minutes. Would be nice if we got 3+ people

https://skribbl.io/?GqCoriIS
I joined, why not
@PaulWhite, @JoshDarnell?
19:51
You all missed out on some fun
Sean is a great artist
LOL I am not, I'm very bad, poor Z had a terrible time guessing
 
2 hours later…
I have finished teaching. Can anyone spare a cigarette.
Guess he found one?
what's happening? The tattooed Brent is next?
Feeling swarthy
@ErikDarling Did it go well? Were people educated?
Did they try to resist at all
22:45
It was a receptive and appreciative audience
@ErikDarling You should go for brisket instead
Melons were twisted
One audience member (probably) said “he rocked my ass and he made me think”
@Forrest I was told there are good restaurants in Grapevine but I’m too lazy to go past the hotel bar
Are you coming to the thing tomorrow?
Absolutely coming to the thing. Looking forward to it. Didn't have the spare energy to submit this year, but happy to hang out with folks.
688 Freeport Pkwy, Coppell, TX 75019 - this place should be relatively close to you and good
What is it called
Oh BBQ
How quaint~~
Looking for anything in particular? I could probably make a recommendation.
22:57
Snooty French always makes me happy
There's a Paris in Texas I think
HtH
Only good snooty French place I know of is in Dallas, sorry.
Where am I
Oh, Dallas is far
Oh, looked it up, might be doable. 19009 Preston Rd #200, Dallas, TX 75252
Good luck. Your other option is la Madeleine, or asking for a baguette at a Subway
Do you have a name for that place? Google maps cannot find it.
23:08
lavendou bistro
Thanks
Hm. About 30 minutes.
I’ll see how adventurous I’m feeling tomorrow

« first day (4994 days earlier)      last day (97 days later) »