I'm considering MariaDB master-master configuration for a webapp database.
My application has some index locking, something like:
SELECT COUNT(*) FROM person WHERE event=? FOR UPDATE;
The transaction checks the subscribers amount to an event and finally insert the new person only if the capacity is not reached, and then the lock is released with the COMMIT command.
I was wondering what happened if I enabled the master-master replica to two servers, for a first test is looks like the lock is completely ignored.
Do you please confirm index/table lock is not working with this configuration?
How usualy is resolved thing kind of problem when someone needs a multi-dbserver environment?
Related
We're trying to move ASP.NET session state for one of our Azure web apps into a database, and it seems like the aspnet_regsql.exe tool is the way to go. Unfortunately, I'm getting stuck on a few issues below. It's an Azure SQL database, and I'm connecting using the server's admin account.
I initially wanted to add the session tables to our existing database, so I ran .\aspnet_regsql.exe -U adminusername -P adminpassword -S servername.database.windows.net -d databasename -ssadd -sstype c. Which throws the exception "Database 'databasename' already exists. Choose a different database name"
Omitting the database name and running it again throws the exception: "Execution Timeout Expired" after about 30 seconds, which is just the default for SqlCommand.CommandTimeout. This occurs while executing the "CREATE DATABASE" command. I tried creating a database manually, and it takes about 50 seconds for some reason. This database is S0 tier and is not under any load
Running aspnet_regsql again on the already-created database (because it's idempotent, right?) leads to the "Database already exists" error, as does pre-creating an empty database for it to start from.
There's no flag that lets me increase the timeout, and I can't set command timeout using the -C (connection string) flag
Adding the -sqlexportonly flag to generate a script and just running that directly doesn't work either (yes, I know I'm not supposed to run InstallSqlState.sql directly). It throws a whole load of error messages saying things like:
Reference to database and/or server name in 'msdb.dbo.sp_add_job' is not supported in this version of SQL Server.
USE statement is not supported to switch between databases.
Which makes me think this script might have some issues with an Azure SQL database...
Does anyone have any ideas?
Update:
It looks like all the errors involving 'msdb' are related to removing and re-adding a database job called 'Job_DeleteExpiredSessions'. Azure SQL doesn't support database jobs, so the only options I can see are
Run SQL on a VM instead (vastly more expensive, and I'd rather stick with the platform services than have to manage VMs)
Implement one of those "Elastic Job Agents"
Perhaps move the same functionality elsewhere (e.g. a stored proc)?
Turns out Microsoft has an article about how to do exactly what I need, which I somehow
missed during my searching yesterday. Hopefully this answer saves someone else a few hours of frustration. All the info you need is at https://azure.microsoft.com/en-au/blog/using-sql-azure-for-session-state/ earlier.
Note that YMMV since it's from 2010 and also says in scary red letters
"Microsoft does not support SQL Session State Management using SQL Azure databases for ASP.net applications"
Nevertheless, they provide a working script that seems to do exactly what I need.
I am trying to debug a performance issue in an ASP.NET application using .NET 4.5, EF5 (with a 2nd level cache and lazy loaded navigation properties) and SQL Server 2014. We are experiencing a number of wait locks in the SQL server. When I look at the locking transactions, they contain a very quick UPDATE, and then a very large SELECT. The UPDATE is ostensibly a necessary one, but I am confused as to why the SELECT is being run in the same transaction (and why anything is being selected at all). The fundamental issue is that the table referenced in the UPDATE statement is locked for the duration of the SELECT statement.
We use repository pattern for getting data from the db, and DbContext.SaveChanges() for committing changes. I cannot figure out how it is possible that EF produces a transaction where there is both a write and a read, and I am not getting relevant results when I try to search Google.
We have a number of interfaces into the system, and a couple of console applications working on the database as well, but they all go through the same setup/versions of .NET and EF.
I figure that it must be through SaveChanges, since this is (AFAIK) the only time that things are written to the database.
Does anyone here have a hint as to how these locking transactions might be produced?
The fundamental issue is that the table referenced in the UPDATE
statement is locked for the duration of the SELECT statement.
The answer is in your question:
the SELECT is being run in the same transaction
X lock is always held until the end of the transaction, i.e. until it commits or rolls back. So if after your quick update there is a long select, all that update locked in your table remains locked until your select ends.
You can separate your update and select if your business rules permit, you can add an appropriate index on the updated table to lock only some rows and not the whole table, or you can optimize your select to execute faster.
Why would the TrackedMessages_Copy_BizTalkMsgBoxDb SQL Agent job start failing with "Query processor could not produce a query plan"?
Query processor could not produce a query plan because of the hints defined in this query. Resubmit the query without specifying any hints and without using SET FORCEPLAN. [SQLSTATE 42000] (Error 8622).
Our SQL guys are talking about amending the stored proc. but we've told them to treat BizTalk db's as a black box
It should go without saying, but before anything, make sure to backup your databases. In fact, if your regular backup jobs are running, you may be able to restore a backup and compare things to when it was working on this server. That said -
Check the SQL Agent Job to make sure no additional steps have been added/no plan has been forced/no hints are being used; it should just have one step called 'Purge' that calls the procedure below with the DB server and DTA database name as parameters.
Check the procedure (BizTalkMsgBoxDb.dbo.bts_CopyTrackedMessagesToDTA) to make sure it hasn't been altered.
If this is a production or otherwise sensitive box, back up the DBs and restore them to a local dev environment before proceeding!
If this is not production, see if you can run the procedure (perhaps in a transaction that you rollback) directly in SSMS. See if you get any better errors. Add print statements to see if you can find out exactly where it's getting conflicting hints.
If the procedure won't run, consider freeing the procedure cache (DBCC FREEPROCCACHE) and seeing if the procedure will run.
If it runs in your dev environment from a backup, you may have to start looking at server/database settings. I can't think of which ones off the top of my head that would cause this error though.
For what it's worth, well intentioned DBAs break BizTalk frequently. They decide that an index is missing or not properly covering, or that security could be improved, or that the database should be treated like other databases they administer are treated. I've seen DBAs do really silly things to the BizTalk databases that get very hard to diagnose.
Did you try updating the statistics on the database table referenced by the stored procedure (which is run by the SQL Server Agent job? The query planner uses those to decide how best to execute your SQL.
As per this how-to, I've successfully configured IIS on my XP-SP3 dev box for SQL Server 2008 Express to save ASP.NET session state information. I'm just using SQL Server because otherwise on every recompile, I was losing the session state which was obnoxious (having to re-login). But, I'm facing an annoying issue in that every time I restart SQL there's this error, and sometimes one or two other very similar friends:
The SELECT permission was denied on the object 'ASPStateTempSessions',
database 'tempdb', schema 'dbo'.
To fix the error, I just open Management Studio and edit the User Mapping for the login/dbo I'm using on the ASPState db, and re-add tempdb to that user with all but deny permissions. Apparently, once the right permissions are there, ASP.NET is able to automatically create the tables it uses. It just can't run that CreateTempTables sproc until the right security is there.
THE QUESTION...
Is there a way to not have to re-do this on every restart of the SQL Server?
I don't really care right now about keeping the temp data across restarts, but I would like to not have to go through this manual step just to get my web app working on localhost, which uses session state variables throughout. I suppose one could resort to some kind of stored procedure within SQL Server to accomplish the task for this machine when the service starts, to not have to do it manually. I'd accept such an answer as a quick fix. But, I'm also assuming there's a better recommended configuration or something. Not seeing an answer to this on the how-to guide or elsewhere here on StackOverflow.
Both answers seem valid; but with most things Microsoft, its all in the setup...
First uninstall the ASPState database by using the command:
aspnet_regsql –ssremove –E -S .
Note:
-E is to indicate you want to use integrated security connection.
-S informs what SQL server and SQL instance to use, and the "." (dot) specifies default local instance
Then re-install using the command:
aspnet_regsql –ssadd –sstype p –E -S .
Note:
The sstype has three options, t | p | c ... the first "t", tells the installer to host all stored procedures in the ASPState database, and all data in the tempdb. The second option "p" tells the installer to persist data to the ASPState database. The last option "c" allows you to specify a different 'custom' database to persist the session state data.
If you reinstall using the "-sstype p" you then need only to supply datareader/datawriter to the ASPState database for the user that's making the connection (in most cases, the application pool's identity in IIS).
The added benefit of persisting the data is that session state is retained even after a restart of the service. The only drawback is that you need to ensure the agent cleanup job is pruning old sessions regularly (it does this by default, every minute).
Important:
If you are running a cluster, you must persist session data. You're only option is to use sstype 'p' or 'c'.
Hope this sheds light on the issue!
For the record, I did find a way to do this.
The issue is that the tempdb is recreated from the model db each time the service restarts. The gist of the solution is to create a stored procedure that does the job, and then make that procedure run at startup.
Source code (credit to the link above) is as follows:
use master
go
-- remove an old version
drop proc AddAppTempDBOwner
go
-- the sp
create proc AddAppTempDBOwner as
declare #sql varchar(200)
select #sql = 'use tempdb' + char(13) + 'exec sp_addrolemember ''db_owner'', ''app'''
exec (#sql)
go
-- add it to the startup
exec sp_procoption 'AddAppTempDBOwner', 'startup', 'true'
go
Well done for finding the strangest way possible to do this.
The correct answer is as follows:
use master
go
EXEC sp_configure 'Cross DB Ownership Chaining', '1'
go
RECONFIGURE
go
EXEC sp_dboption 'ASPState', 'db chaining', 'true'
go
I have a Web site live and running now. I am using the Subsonic to handle the database connections etc.
I am getting time out expired error while updating a table (say Employee). When I check sp_who2, I see the suspended connection for the PID which is updating with a block by anothor pid, so I run the profiler and found out when ever this suspended connection occur, the blocked pid is a select statement on the view (say ActiveEmployees, which is the same as the table but with some where conditions).
Anyone know why a Select statement on the view could cause failure in update. If it is other (like select fails due to update) may be reasonable.
Is there any way for me to make select on a view without locking the table?
PS: I am using the Sql server 2005 and subsonic 2.2.
You might add with(nolock) hint to the select statement in the view if you don't care about accuracy of the returned data (it will return uncommited rows possibly).
We encountered timeouts also when the select statements where scanning a table that other thread was inserting into. I resolved the issue by adding appropriate index that is used by our select.