A log backup is not based on the most recent full backup. It is based on the most recent log backup.

I.e., you can "skip" a full or differential backup when you restore log backups, as long as you have an unbroken chain of log backups.

As for your particular issue: Work with the backup software/vendor and see if you can make it perform the snapshot backups so they are seen as COPY_ONLY by SQL Server. Possibly a setting in your backup software.

I have one SQL Server Standard (2019), I want to add one server to developper (not used by production so the version if developper), it's possible to replicate data ? what is the best methode ? we want to test one batch with data production.

Thanks

Thanks.

I want one way to have a server with production data only to test our batch. Replication between standard and developper editions it's possible (technologically) ? For licenses I think it's not problem because only batch for test will connect this. I have never use replication in sql server, I have 2019 standard, what is the best method ? Thanks

The https://docs.microsoft.com/en-us/sql/t-sql/statements/grant-database-permissions-transact-sql?view=sql-server-ver15 says "ADMINISTER DATABASE BULK OPERATIONS Applies to: SQL Database."

In the context of the docs "SQL Database" is Azure SQL DB. Not the SQL Server running on Windows/Linux. "SQL Database" always means Azure SQL Database. The docs simply use both interchangeably.

after alot of reading i found multiple sources saying postgres should be able to handle ~100 million rows

i set up a simple table with:

CREATE TABLE trade
(
    id INT PRIMARY KEY NOT NULL,
    symbol VARCHAR(),
    open_time TIMESTAMP,
    end_price INT
)

i have 12 million of thies records

my simple query of

SELECT * FROM table WHERE symbol=x and open_time>start_time and open_time

this query always returns less than a 1000 rows yet it take 1100 millisecond this seams like alot for a simple table with 10x less rows than it should handle?

altho i dont have any indexes cause i dont know what would be the best indexs to put on this table.

is it possible to get this down to 100ms?

any help writing a more performant query or db would be appriciated

EDIT 1:

after reading the comments i put the following indexes on the table

CREATE INDEX open_time ON trade(open_time);
CREATE INDEX symbol ON trade(symbol);
CREATE INDEX end_price ON trade(open_price);

after adding this the query time is 240 ms is this the max?

altho i have noteced querying the beginging 40k rows the query time drops to 60ms
after that is rises to 240ms what is causing this?

I have a table that tracks the action of my users (used for logging and debugging for their purpose).

From time to time, I need to mark the data as treated by doing the following call :

UPDATE actions
SET is_treated = 1
WHERE code = 200
    AND is_treated = 0
    AND account_uuid = :uuid
    AND DATE(executed) = UTC_DATE()
LIMIT 1000

I do that command inside a loop that loop as long as there are "is_treated" at 0 for that day. I limited to 1000 to fasten the query.

That table currently have around 1M entries.

Despite setting the limit at 1000, the query fails from time to time with...

Lock wait timeout exceeded; try restarting transaction

I tried reducing the limit to 100, with the same error.

For information, the table as the following indexes :

$> SHOW INDEX FROM actions;
+-------------+------------+---------------------------+--------------+--------------+-----------+-------------+----------+--------+------+------------+---------+---------------+

| Table       | Non_unique | Key_name                  | Seq_in_index | Column_name  | Collation | Cardinality | Sub_part | Packed | Null | Index_type | Comment | Index_comment |

+-------------+------------+---------------------------+--------------+--------------+-----------+-------------+----------+--------+------+------------+---------+---------------+

| actions |          0 | PRIMARY                   |            1 | id           | A         |      952794 |     NULL | NULL   |      | BTREE      |         |               |

| actions |          1 | account_uuid              |            1 | account_uuid | A         |         696 |     NULL | NULL   | YES  | BTREE      |         |               |

| actions |          1 | api_token                 |            1 | api_token    | A         |         848 |     NULL | NULL   | YES  | BTREE      |         |               |

| actions |          1 | ix_actions_code       |            1 | code         | A         |           6 |     NULL | NULL   |      | BTREE      |         |               |

| actions |          1 | ix_actions_executed   |            1 | executed     | A         |      952794 |     NULL | NULL   |      | BTREE      |         |               |

| actions |          1 | ix_actions_is_treated |            1 | is_treated   | A         |           2 |     NULL | NULL   |      | BTREE      |         |               |

+-------------+------------+---------------------------+--------------+--------------+-----------+-------------+----------+--------+------+------------+---------+---------------+

Is there a way to optimize the query to be run processed faster ?

I was wondering if doing something like the follow would not be faster.

UPDATE actions
SET is_treated = 1
WHERE id IN
(
    SELECT id
    FROM actions
    WHERE code = 200
        AND is_treated = 0
        AND account_uuid = :uuid
        AND DATE(executed) = UTC_DATE()
    LIMIT 1000
);

What do you recommend?

Thank you in advance.

Trying to view all executions of a named stored procedure where execution time exceeds a given number of seconds. How do you filter on execution time (seconds)? This is what I have cobbled together so far

CREATE EVENT SESSION [Slow SP Executions]
ON SERVER
    ADD EVENT sqlserver.module_end
    (SET collect_statement = (1)
     ACTION
     (
         sqlserver.host,                       
         sqlserver.database_name,  
         sqlserver.client_app_name, 
         sqlserver.session_server_principal_name, 
         sqlserver.username,                      
         sqlserver.sql_text,
         sqlserver.tsql_stack
     )
     WHERE (
               [object_type] = 'P ' 
               AND [sqlserver].[database_name] = N'MyDB'
               AND [object_name] = N'MySproc'              
           )
    )

In WSFC FCI with 2 nodes and file share, each node participates in a heart beat.

The file share is a voting witness, and connectivity from any node to that share is counted as an affirmative vote. More than one-half of the possible votes must be affirmative for the cluster to be healthy.

So this mean - in a 2 node fci with 1 file share, each node sends its heart beat to the cluster and its connectivity to witness is counted as 1 vote. So each node effectively sends 2 votes to the cluster?

I'm trying to get drive space for multiple servers

$servers = Get-Content "\\SQL-Server1\e$\Scripts\DBInfo\DBInfo_DiskfreeComputers.txt"
foreach ($Server in $Servers)
{
$results = Get-CimInstance -ClassName Win32_LogicalDisk

$DriveLetter = $results.DeviceID

$FreeSpace = $results.FreeSpace

$TotalSpace = $results.Size

$CollectionDate = $([System.Datetime]::Now.ToString("dd/MM/yy HH:mm:ss"))

Write-Host "Insert into SQLSERVER.dbo.DBInfo_DiskSize(ServerName, DiskLetter, FreeSpace,TotalSpace,CollectionDate)

Values $Server, $DriveLetter, $FreeSpace, $TotalSpace, $CollectionDate"

}

This is giving me:

Insert into SQLSERVER.dbo.DBInfo_DiskSize(ServerName, DiskLetter, FreeSpace,TotalSpace,CollectionDate)
            Values SQL-SERVER, C: E: F: Z:, 13104119808 244906799104  16500219904, 214260355072 297121345536  246957469696, 04/05/22 16:00:48

How would I get this to format to do an insert for each disk, instead of all on this one statement - for instance this server has 4 disks I'd like to insert 4 rows into the table one per disk

I am new to MSSQL. I have a question about executing MSSQL Stored Procedure.

Let me briefly talk about my development environment first:

• PC400 (Computer/Server A), where SQL Server 2012 is installed
• PC401 (Computer/Server B), where Java application is running on

PC400 has SSIS project & package(s) deployed to its SSISDB. Of course, I have no problem in executing the packages LOCALLY in PC400.

But this is not what I need. I want the java application installed on PC401 to be able to execute those packages stored in PC400. I have no problem in using a sa/temp user account with "SQL Server Authentication" to login the server:

But based on my research, "SQL Server Authentication" does not allow me to deploy/execute packages in SSISDB.

Then, I found the following advice: https://dba.stackexchange.com/questions/39614/connect-to-sql-server-using-windows-authentication-from-another-pc-without-activ

Let say the windows user account of PC401 is "HKB\Hello123". By creating an identical Windows user under "MSSQL -> Security -> Logins" in PC400:

my java program is able to use the following codes to execute a stored procedure in PC400 using "Windows Authentication":

con = DriverManager.getConnection("jdbc:sqlserver://HKA-PC400:1433;DatabaseName=TempTest;integratedSecurity=true"); 
CallableStatement cs = null;
cs = this.con.prepareCall("{call SP_ETL_B}");
cs.execute();

But is there any other methods to achieve this? Besides, I hope my java program would be able to use a sa/temp user account("SQL Server Authentication") to pretend to be a "Windows Authentication" one OR simply connect to an existing Windows user account, then to trigger Stored proc/SQL Server Agent job to run the packages in PC400.

I need to be able to estimate server resources for sql server To be able to save money. For example, how many resources (CPU, memory and RAM) are needed for a database with 3,000 users. Is there a formula? Is it possible to respond to this number of users with one server?