meta user
network profile
| bio | website | |
|---|---|---|
| location | ||
| age | ||
| visits | member for | 10 months |
| seen | Jan 3 at 11:42 | |
| stats | profile views | 3 |
|
Aug 22 |
comment |
What units are 'reads' and 'writes' in dm_exec_query_stats Thanks Aaron, I had a suspicion it would be pages. I marked your answer as correct because of your references. |
|
Aug 14 |
comment |
Make sure two columns are sorted in the same order I don't think you can guarantee uniqueness with just CreateDateTime. Clocks can be reset forward and backwards. Events can happen so quickly they appear to happen at the same time. (I almost convinced myself I could make a datetime2 column the PK of my log table, but you really need to include an IDENTITY column to ensure uniqueness). |
|
Aug 14 |
comment |
Make sure two columns are sorted in the same order Also be careful your GETDATE() doesn't jump backwards or forwards in time (eg: because someone adjusts the server clock, or daylight saving). |
|
Aug 3 |
comment |
single select vs multiple selects using limit and offset According to TechNet Read Committed Snapshot and Snapshot isolation levels are enabled by default in Azure; query sys.databases to confirm. Both those isolation levels allow selects to read data even if records are locked by another transaction (even Serializable ones). Your reader will see an older version of the data. |
|
Jul 20 |
comment |
Optimising MySQL @randymelder - I stand corrected! |
|
Jul 19 |
comment |
Sql server 2008 filestream maximum number of rows I've been toying around with filling NTFS folders with obscene numbers of files recently (random pet project of mine). Performance tanks in Explorer around 1M, but most file system operations are OK up until you exceed the file system cache. 20k files shouldn't pose a problem. |
|
Jul 16 |
comment |
Poor performance after changing SQL Server 2008 R2 databasefinagle verb - to improve the performance of a SQL query by tweaking an incorrectly chosen JOIN operator. Eg: "My query suddenly ran slow, so I finagled the merge join into a hash join" |
