Looking forward to attend SQL Saturday Atlanta.

Finally get around to read @straczynski.bsky.social excelent The Glass Box. I can highly recommend this near future story that felt all to real. Note to my self, Kindle UK as not the same as Kindle US sot got the book two month after it was relased.

Looking forward to see Murderbot on TV

MWmarthawells.bsky.socialMar 20, 2024 2:24pm

More TV news: moviesr.net/p-sci-fi-ser...

Sci-Fi Series ‘Murderbot’ Officially Begins Filming at Apple TV+

Based on the Martha Wells novel series, The Murderbot Diaries, the series is led by Alexander Skarsgård, who plays the self-hacking security android horrified by human emotion.

An idea worth voting for! SSMS Right Click To Open Execution Plans In A New Tab: feedback.azure.com/d365communit...

Great, good to have an explanation. Guess it make sence to have FK validates against the current state and not the version store. Thanks you!

Good question. Yes, table does both have a foreign key and another table have a foreogn key on the table.Good question. Yes, table does both have a foreign key and another table have a foreogn key on the table.

Sure, what ias throwing me of is that is that the dead locks are between S locks and X locks. All Key locks by the way. Seems like MERGE is not using the version store to read the data but instead tried to take a shared lock.Sure, what ias throwing me of is that is that the dead locks are between S locks and X locks. All Key locks by the way. Seems like MERGE is not using the version store to read the data but instead tried to take a shared lock.

Not sure but asume it is a real dead lock as I get the dead lock alert inlcuding dead lock graph in Redgate SQL Monitor.Not sure but asume it is a real dead lock as I get the dead lock alert inlcuding dead lock graph in Redgate SQL Monitor.

I did already find you block and seems to be similar but as you say without the RCSI part. There are some temporary table DDL included as the merge are between the temp table and the target table. The target table do have one non-clustered index but the dead lock is on the PK index.I did already find you block and seems to be similar but as you say without the RCSI part. There are some temporary table DDL included as the merge are between the temp table and the target table. The target table do have one non-clustered index but the dead lock is on the PK index.

Any ideas how two sessions running the same MERGE stagement can dead lock each other when usign RCSI? #sqlhelp Yes, I know you shouldn't use merge but you ahve what you have.