Twitter | Search | |
Paul Randal
Owns SQLskills, wrote CHECKDB+DBCC, managed Storage Engine, internals/waits/DR expert, Master Diver, fisherman, bibliophile. Married to .
35,832
Tweets
15
Following
15,001
Followers
Tweets
Paul Randal 17h
Replying to @NedOtter @jdanton
No - pages are always compressed. There is never a cached copy of any decompressed data. It’s decompressed on demand every time it’s accessed.
Reply Retweet Like
Paul Randal 21h
But the ones I made for were *very* different from what others asked for...
Reply Retweet Like
Paul Randal 21h
Replying to @NedOtter @jdanton
very common to see high OLEDB from monitoring tools running DMVs in loops, or CHECKDBs. Possibly linked servers, but most likely a monitoring tool.
Reply Retweet Like
Paul Randal 21h
Replying to @NedOtter @jdanton
Not surprised. SOS_SCHEDULER_YIELD because you've got lots of data in memory and the update stats is burning CPU decompressing it all. LATCH_EX probably an ACCESS_METHODS latch from parallel scans. OLEDB - could be all kinds of things.
Reply Retweet Like
Paul Randal 21h
Replying to @SQLJuJu @BrentO and 2 others
Not sure what I did, but sure do appreciate the nice words :-) Thanks ! Community is where it's at!
Reply Retweet Like
Paul Randal Feb 26
[Blog] The Curious Case of… whether corruption can propagate to secondary databases
Reply Retweet Like
Paul Randal Feb 22
It was all 's fault... intolerance... :-)
Reply Retweet Like
Paul Randal Feb 21
Replying to @randy_knight
In that case it seems like a bug state where it's not properly removing it from the AG when you set HADR to OFF. You could try blipping the database offline and online to completely remove any leftover status.
Reply Retweet Like
Paul Randal Feb 21
Replying to @randy_knight
Could be case similar to this:
Reply Retweet Like
Paul Randal Feb 21
Replying to @randy_knight
What percent full is the log?
Reply Retweet Like
Paul Randal Feb 21
Replying to @randy_knight
Have you done a checkpoint since setting to simple? LRW_desc is why it couldn't clear last time it tried, not the status of the log 'right now'.
Reply Retweet Like
Paul Randal Feb 20
Got into 3-d printing in a big way recently, as did . Latest: swept triangle from . 20hr print at 0.15mm, 60% speed, in shiny blue TTY3D PLA on a Prusa i3 Mk2S. Just needs a little sanding - very pleased!
Reply Retweet Like
Paul Randal Feb 19
Replying to @A_Bansal
1070, and 2017 is 931
Reply Retweet Like
Paul Randal Feb 18
Replying to @SQLPoolBoy @chuckh1958
Yup - reorg and shrink both set themselves to be the deadlock victim and that can't be changed. Due to the weird nature of some of the locking they do.
Reply Retweet Like
Paul Randal Feb 16
Happy to say my precon at is sold out! Looking forward to being there for the first time :-)
Reply Retweet Like
Paul Randal Feb 14
Replying to @tradney
Tell him to watch out for school buses ;-D
Reply Retweet Like
Paul Randal Feb 13
Replying to @ecreart
Always email, which you did - thanks
Reply Retweet Like
Paul Randal Feb 10
We haven't been - not a big fan of shore diving, especially with 's camera rig.
Reply Retweet Like
Paul Randal Feb 10
Reply Retweet Like
Paul Randal Feb 9
[Blog] The Curious Case of… restoring a table into a different database
Reply Retweet Like