Thank you Scott.

Thanks everyone.

Link to estimated plan

https://www.brentozar.com/pastetheplan/?id=rktPkssh2

Hi Johan, Thank you. Actually, I wanted to know the reason the difference between the two methods.

Thanks a ton Phil.

Thank you everyone. It turned out to be a network issue.

• This reply was modified 1 year, 4 months ago by  vsamantha35.

Thanks a lot Johan and Jeffrey.

One thing, I wanted to ask. Is there a way to run deletes in small batches to minimize blocking especially when child tables are involved...

just created a login and tried it. It gave some good layman explaination. Once thing, I wanted to ask, why...

From the nature of some of the code, it looks like all parameters are being passed as NVARCHAR(4000).

It doesn't matter how well code may be written, that generally means...

This isn't an...

Any thoughts on waittypes?

without even knowing which CRM vendor this is I would suggest the following

but your locking is typical of readers blocking writers (which RCSI "solves") - and fact that that...

Created a table in such a way, each rowsize is one 8K page.

Now when I execute this

SELECT

T.name as 'Table',

ix.index_id,

ix.[name] AS IndexName,

ix.type_desc AS IndexType,

[ips].[index_level],

ips.page_count AS [PageCount]

FROM sys.dm_db_index_physical_stats (DB_ID(), NULL, NULL, NULL,...

Is there a way what queries caused the data growth and what tables grown big during that time frame? Any ideas?

Data movement will be always there as it is synchronous commit replica and this logic works for doing index rebuilds without any issue. But for update stats why it is...