Postgres 13 cluster in Debian Linux server contains 30 databases. Databases contain number of schemas.
How to find biggest files which occupy most space in disk ?
I tried
select
relname::char(25),
pg_size_pretty(pg_total_relation_size(c.oid))::char(10) as totalsize,
n.nspname::char(12),
case
when c.relkind='i' then 'index'
when c.relkind='t' then 'toast'
when c.relkind='r' then 'table'
when c.relkind='v' then 'view'
when c.relkind='c' then 'composite type'
when c.relkind='S' then 'sequence'
else c.relkind::text
end ::char(14) as "type"
from
pg_class c
left join pg_namespace n on n.oid = c.relnamespace
left join pg_tablespace t on t.oid = c.reltablespace
where
(pg_total_relation_size(c.oid)>>21)>0 order by
pg_total_relation_size(c.oid) desc
But it returns sizes for current database only. How to run in over whole cluster ? Can some plpgsql script used for this.
Output should include database name column.
Client application uses psqlODBC to get data so psql or shell scripts should preferably avoided.
You cannot do that, as you can only query the database to which you are connected. You need to connect to each database in turn.
Related
SELECT DISTINCT
OBJECT_SCHEMA_NAME (sc.object_id) as "schema", OBJECT_NAME(sc.object_id) as "name", sc.*
-- FROM syscomments sc
FROM sys.sql_modules sc
WHERE "Definition" LIKE '%raiserror%'
and
OBJECTPROPERTY(object_id, 'IsMSShipped') = 0
and
OBJECT_NAME(sc.object_id) like '%diagram%'
Why is this query returning these SPs? Aren't they from Microsoft?
sp_helpdiagramdefinition
sp_creatediagram
sp_renamediagram
sp_alterdiagram
sp_dropdiagram
IsMSShipped is set to 1 for any object that was created during SQL Server's installation. The Diagram objects are optional and are only added to a database after the initial installation.
In other words, although they are from MS, they are not Shipped from MS (at least not as MS is defining "Shipped").
Yes I know, it's dumb, everyone gets tripped up by this at least once. They should have called it something like IsMSInstalled instead. Just goes to show the importance of picking good names.
The SOP way to handle this is to also filter on the schema ("sys" is always schema_id 4).
I have a list of jobs to run multiple ssis packages. I have a big list, I want to know if it is possible to run a tsql to msdb and get the information of the packages in the steps with the connection strings.
Super old thread but the below code is slightly adapted and much more useful as it shows package level details.
select
prj.name as 'ProjectName'
,pa.name as 'SSISPackageName'
,op.parameter_name as 'ParmaterName'
,op.design_default_value as 'ConnectionString'
from
catalog.object_parameters op
join catalog.projects prj
on op.project_id = prj.project_id
join [catalog].[packages] pa
on pa.project_id = prj.project_id
where op.parameter_name like '%ConnectionString%'
If your packages are stored in the SSISDB Catalog, then you would need to query on SSIDB database to get the actual connection strings. Here is what I would suggest, use the query that Rodrigo A has provided and tweak the column 'Command' further to get the packages name from it by using string functions and have the Step ID and Package Name as output. use the list of packages that you have obtained and put it in the IN clause of the following query and run it against SSISDB -
select prj.name as 'ProjectName'
,op.object_name as 'SSISPackageName'
,op.parameter_name as 'ParmaterName'
,op.design_default_value as 'ConnectionString'
from catalog.object_parameters op
join catalog.projects prj
on op.project_id = prj.project_id
where op.parameter_name like '%ConnectionString%'
and op.object_name in (
--put the package name list here
);
You can then join on the result sets, to get the step Id, PackageName and the connection strings in that package.
Run this query over the server you want to check:
SELECT [sJSTP].[step_id], Name, sJSTP.Command FROM [msdb].[dbo].[sysjobs] [sJOB]
LEFT JOIN [msdb].[dbo].[sysjobsteps] AS [sJSTP]
ON [sJOB].[job_id] = [sJSTP].[job_id]
AND [sJOB].[start_step_id] = [sJSTP].[step_id]
ORDER BY Name, step_id
Cheers!
How to find recently created index details in my SQL Server database? Any query to find this?
In my database there are a lot of indexes. I want to know which of those indexes were recently created, with all their details.
You can use SCHEMA changes history to know index creation changes along with many changes
Below is how you do it..
1.Right click server
2.Goto reports -->standard reports-->Schema changes history
below is screenshot from mt device
Default trace is enabled by default,unless you turn it on
below query tells you,if default trace status is ON
select * from sys.configurations where name like '%trace%'
below query can provide object creation stats
SELECT OBJECT_NAME(objectid),objectname,indexid
FROM sys.fn_trace_gettable(CONVERT(VARCHAR(150), ( SELECT TOP 1
f.[value]
FROM sys.fn_trace_getinfo(NULL) f
WHERE f.property = 2
)), DEFAULT) T
JOIN sys.trace_events TE ON T.EventClass = TE.trace_event_id
where DatabaseName=db_name()
ORDER BY t.StartTime ;
We license our software by number of workstations.
I have a query that I have used for years to get an accurate count of the workstations logged in to my SQL Server database. For simplicity, all users use the same login name/password. This is built in to the script that attaches to the DB. They have access only to that DB with the exception of
USE [Master] GRANT VIEW SERVER STATE to MyUser
The query that has been working is below:
SELECT COUNT(Users) AS UserCount FROM (SELECT COUNT(Master.dbo.sysprocesses.hostname) AS Users FROM Master.dbo.sysprocesses LEFT OUTER JOIN Master.dbo.sysdatabases ON Master.dbo.sysdatabases.dbid = Master.dbo.sysprocesses.dbid WHERE (Master.dbo.sysdatabases.name = 'MyDatabase') GROUP BY Master.dbo.sysprocesses.net_address) AS UserCount_1
Basically this relies on the mac address (Master.dbo.sysprocesses.net_address), since both Workstation names and ip addresses can be duplicated.
Recently this has stopped working at a number of customers. Suddenly individual workstations are showing multiple net addresses for the same workstation causing a substantial overcount of users. This may be related to SQL Server 2012 - not sure.
What I need is a very reliable way to get a count of workstations logged in to my database.
If anyone can tell me why I am suddenly getting multiple net_addresses for each workstation and how to prevent that that would be one possible solution.
Otherwise if anyone can give me a rock solid way to get a workstation count other than the above that would be great. Our largest customer is 50 users by the way.
HERE IS AN EXAMPLE:
SELECT Master.dbo.sysprocesses.hostname AS Users, Master.dbo.sysprocesses.net_address FROM Master.dbo.sysprocesses
LEFT OUTER JOIN Master.dbo.sysdatabases ON Master.dbo.sysdatabases.dbid = Master.dbo.sysprocesses.dbid
WHERE Master.dbo.sysdatabases.name = 'mydb' GROUP BY Master.dbo.sysprocesses.hostname, Master.dbo.sysprocesses.net_address
RETURNS:
DAVID-PC 001CC490239E
FLOOR1 001CC41D8012
FLOOR2 CB8FEE6C5856
FLOOR3 A50B18FF1516
KER-PC7 6C626DEA68CC
LIZ-PC A4E553460E35
LIZ-PC EFE3F0E20260
LIZ-PC FD32F7B30360
PAP 9D35A704C29C
PAP CFB724BA1183
PAP D1A58A8878E6
PAP E9B116CA34B8
PAP F38B335A7AE6
Thanks in advance for any help.
You can get an accurate count of the users logged in to SQL Server database by the following query:
SELECT
DB_NAME(dbid) as DB
COUNT(dbid) as Numb
loginame as LoginNa
FROM
sys.sysprocesses
WHERE
dbid > 0
GROUP BY
dbid, loginame
Also you can get full details of connected users by this:
sp_who2 'Active'
Here is a way that will give you the number of IPs or MAC addresses connected to a given database. I'd go with IP as you might have multiple NICs/MACs in a given high end workstation.
SELECT COUNT(DISTINCT c.client_net_address) as DistinctIPCount,
COUNT(DISTINCT p.net_address) as DistinctMACCount
FROM sys.dm_exec_connections c INNER JOIN sys.dm_exec_sessions s ON c.session_id = s.session_id
INNER JOIN sysprocesses p ON s.session_id = p.spid
WHERE s.is_user_process = 1
AND p.dbid = DB_ID('yourdbnamehere')
I have some offline databases on a SQL server. I would like to know which files on disc are related to these databases. Is it possible to retrieve the file list of offline databases without taking them online first?
This will give you a list of all physical file paths related to any offline databases, along with database name and file type:
SELECT
'DB_NAME' = db.name,
'FILE_NAME' = mf.name,
'FILE_TYPE' = mf.type_desc,
'FILE_PATH' = mf.physical_name
FROM
sys.databases db
INNER JOIN sys.master_files mf
ON db.database_id = mf.database_id
WHERE
db.state = 6 -- OFFLINE
Or simply
select * from sys.databases where state_desc='OFFLINE'
List all the available, but offline SQL server database files
The following statement will list all the files associated with offline SQL server databases
SELECT
m.physical_name + '\' + m.name AS [file_path]
FROM
sys.databases AS d
INNER JOIN sys.master_files AS m ON d.database_id = m.database_id
WHERE
d.state_desc = 'OFFLINE'
--AND m.type_desc = 'ROWS'
GROUP BY
m.physical_name + '\' + m.name
Note: Uncomment the line AND m.type_desc = 'ROWS' (delete the --) to filter the list further to include only database files. Otherwise, log files will also be listed.
The GROUP BY clause is there to prevent entries appearing more than once.