Forum Replies Created

Viewing 15 posts - 8,776 through 8,790 (of 59,078 total)

  • Reply To: deadlock issue

    Actually, the code is doing one UPDATE and one DELETE.

    The UPDATE needs an index on LOCK_GROUP_STR  and ROWID_TABLE to make it fast.

    The DELETE needs an index on just the LOCK_GROUP_STR...

  • Reply To: using trigger for tracing

    ScottPletcher wrote:

    Of course bad triggers can be written.  Bad stored procedures can also be written, but that doesn't mean we should quit writing stored procedures (agreed?).   Yes, take great care...

  • Reply To: Index Usage

    I'll have to say "oh, be careful now" and yell "It Depends"!

    The presence of only "Seeks" in the usage stats must not be interpreted as something meaning "everything is great". ...

  • Reply To: Advice for a Rookie DBA

    Sopheeyah wrote:

    Also, what does it entail to be very successful in this field?

    Tenacity, lots and lots of study, lots and lots of practice, and a whole lot of real life...

  • Reply To: ColumnStore Indexes not releasing the Deleted Rows

    TRACEY-320982 wrote:

    Once you rebuild all partitions how do you find which partition to build, as I would think I don't want to rebuild all partitions every time as that would...

  • Reply To: The First Rule

    RonKyle wrote:

    You go ahead and do your de duplication that will inevitably occur with some surrogate only keys.  I've seen these results many times.  I have yet to see a...

  • Reply To: The First Rule

    RonKyle wrote:

    This is one reason I dislike many natural keys as PKs.

    I don't understand how this follows from the discussion..  First, there's not a whole lot of cases where there...

  • Reply To: Data Migration

    Phil Parkin wrote:

    I'm just waiting for Jeff to chime in on this one ...

    I've not done an Oracle to SQL Server migration since 2002 and that was a "one time" transfer.

    Since...

  • Reply To: Export to CSV

    Jenny77 wrote:

    Hi,

    i am trying to export a table into csv, and it should be comma delimited with doublequotes around each value like this:

    "value1","value2","value3"

    this works fine, but one of the columns...

  • Reply To: The First Rule

    Knuth once wrote the parable of "Premature optimization is the root of all evil".  That has been interpreted (actually, improperly warped) by many as leaving in place NUMERIC(18,0) for integers...

  • Reply To: A DateRange Table Valued function

    Nice function, Jonathan.  Well done!

  • Reply To: Number of Mondays between two dates

    Sergiy wrote:

    Look at the progress we've got in SQL over the years!

    what used to be 13 years ago a short inline formula using system functions and integer arithmetic is now...

  • Reply To: Primary Key and Clustered Index

    RonMexico wrote:

    Thanks for all of the information. It was very helpful because some things I read swear by adding an identity column for the primary key regardless of what type...

  • Reply To: Give Up on Natural Primary Keys

    hmbacon wrote:

    OH is actually an alfa surrogate for Ohio.

    Not by definition.  Part of the definition of a surrogate key is that you should be able to derive NO information by...

  • Reply To: Give Up on Natural Primary Keys

    nova wrote:

    Don't get distracted by the arbitrary idea that one key or other has to be primary key. In practical terms it doesn't matter which of the keys in a...

Viewing 15 posts - 8,776 through 8,790 (of 59,078 total)