August 5, 2026 at 6:08 pm
Was? Is? meaning you tried it? If this is your post int he test thread, I see your response with 3 links.
August 5, 2026 at 6:39 pm
my post had 4 links when I created it.
I have now edited it and added another 3 links (which also failed before if we added a second link to any post
August 6, 2026 at 2:33 pm
Customer: We've been told index maintenance is taking too long because we have hundreds of millions of records in some tables on this VLDB. Data is ITD, so cannot be archived. Can you partition?
Me: Sure. Let me take a look.
First table, 27 columns, 15 indexes and only one column isn't listed in any of those indexes.
Who has the pork chop launcher?
August 6, 2026 at 4:43 pm
I don't know how many of you have seen this, but OUCH. True, funny, painful, true.
August 8, 2026 at 7:21 pm
I tried some things over on that thread where the post-spam-fix fix was being tested. I tried to break it like before and WOO-HOO!!! It looks like we're back in business!!!.
--Jeff Moden
Change is inevitable... Change for the better is not.
August 10, 2026 at 10:16 am
I don't know how many of you have seen this, but OUCH. True, funny, painful, true. https://www.stvn.sh/writing/programming-still-sucks-fqffhyp
That was far too close to reality.
And the doll is on fire.
"The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood"
- Theodore Roosevelt
Author of:
SQL Server Execution Plans
SQL Server Query Performance Tuning
August 13, 2026 at 11:40 am
SPAM has officially spread to the messages queue. In the past 24 hours, I've gotten 3 messages from the same user that look very spammy.
gsgd
Account
Last login: August 13th 2026
Login count: 5
Reputation: Newbie
Points: 5
August 15, 2026 at 12:07 pm
SPAM has officially spread to the messages queue. In the past 24 hours, I've gotten 3 messages from the same user that look very spammy. gsgdAccountLast login: August 13th 2026Login count: 5Reputation: NewbiePoints: 5
I think they've relaxed a bit on spam because the anti-spam rules were keeping people from posting. For example, me. It's not uncommon for me to answer a post with more than 1 link and it's not uncommon for me to rapidly post more than one reply on any given thread to answer more than one person on the thread.
Getting spam on your private messages on this site is another thing, though. I've never seen that before.
--Jeff Moden
Change is inevitable... Change for the better is not.
August 15, 2026 at 12:24 pm
Customer: We've been told index maintenance is taking too long because we have hundreds of millions of records in some tables on this VLDB. Data is ITD, so cannot be archived. Can you partition? Me: Sure. Let me take a look.
First table, 27 columns, 15 indexes and only one column isn't listed in any of those indexes.
Who has the pork chop launcher?
I'm working with a 3rd party vendor that someone hired without checking with the DBA team that create a product that uses a table with 712 columns. The columns are named after the datatype and which occurrence number the data type is. For some examples..
VARCHAR50_9
VARCHARMAX_7
INT_12
DATETIME_15
For us, most of the columns are null. And, they use SSAS with a ton of views on the table to "customize" their train wreck for us.
Their archive process created a quarter Tera Byte log file and they weren't even aware. Their latest archive process is to delete the latest 3 million rows and replace them with the latest 3 million rows from the source. And, yes... it is actually a DELETE.
And, no... I couldn't make this up if I tried.
And, I almost forgot... they decided to do all of this, including the SSAS stuff, as an afterthought... and they did it on SQL Server Standard Edition and wonder why they're having memory problems, especially with SSAS.
--Jeff Moden
Change is inevitable... Change for the better is not.
August 28, 2026 at 1:56 am
Just testing after a fix on Norton Web Safe on my laptop.
Another fine reminder that "Change is inevitable... change for the better is not".
--Jeff Moden
Change is inevitable... Change for the better is not.
August 28, 2026 at 1:58 am
And... that worked.
--Jeff Moden
Change is inevitable... Change for the better is not.
August 28, 2026 at 8:26 am
That's an original way to handle your column names. Pretty much need a metadata mapper to translate your queries. Hope they don't change their datatypes often
August 28, 2026 at 3:36 pm
Glad it worked. I assume a Norton update?
I find more and more people changing datatypes. I think I might need to adjust my "Zero downtime" talk to include that.
September 7, 2026 at 4:50 pm
Glad it worked. I assume a Norton update?
I don't know if it was a Norton Update or not because I've not heard of anyone else having such a problem with the Norton product.
I find more and more people changing datatypes. I think I might need to adjust my "Zero downtime" talk to include that.
I'm guess you're responding to the 712 column "wonder" that I posted?
On that note, this is why a DBA needs to be brought in early on a design even if there aren't such problems as what I posted there. I've worked on teams where the Front-End programmers were the ones that designed the tables. Some do a pretty good job. Some let the machine do things for them and you end up with NUMERIC(19) for just about everything even for TinyInt or Bit requirements and NVARCHAR(256) for a whole lot of columns that might only need a CHAR(5) for example. Somehow, that type of thing makes it to production and when it gets hit with a little scale, people suddenly "get the datatype religion" but you can't afford the down time and, and, and.
On LinkedIn, a whole lot of people are preaching the "Done > Perfect" idea... And they have no idea what "Done" actually means. They frequently mean "shipped = 'Done'" and I don't need to tell you what can go wrong there. 😉
--Jeff Moden
Change is inevitable... Change for the better is not.
September 8, 2026 at 12:22 pm
Steve Jones - SSC Editor wrote:Glad it worked. I assume a Norton update?
I don't know if it was a Norton Update or not because I've not heard of anyone else having such a problem with the Norton product.
Steve Jones - SSC Editor wrote:I find more and more people changing datatypes. I think I might need to adjust my "Zero downtime" talk to include that.
I'm guess you're responding to the 712 column "wonder" that I posted? On that note, this is why a DBA needs to be brought in early on a design even if there aren't such problems as what I posted there. I've worked on teams where the Front-End programmers were the ones that designed the tables. Some do a pretty good job. Some let the machine do things for them and you end up with NUMERIC(19) for just about everything even for TinyInt or Bit requirements and NVARCHAR(256) for a whole lot of columns that might only need a CHAR(5) for example. Somehow, that type of thing makes it to production and when it gets hit with a little scale, people suddenly "get the datatype religion" but you can't afford the down time and, and, and. On LinkedIn, a whole lot of people are preaching the "Done > Perfect" idea... And they have no idea what "Done" actually means. They frequently mean "shipped = 'Done'" and I don't need to tell you what can go wrong there. ??
If you want real fun regarding "shipped = 'Done'", research Outcome-Based Delivery. That's the new thing.
Well, I say "new" but we've been doing it for almost 4 years at this point.
Viewing 15 posts - 66,916 through 66,930 (of 66,931 total)
You must be logged in to reply to this topic. Login to reply