I will have a variety of different types of work that will come into my Service Broker queue and I'll likely have a stored procedure or two for each of the different types of work (ie. move order header, move items, move payment, etc.) What is required to be done in each of these steps may vary by the subsidiary and type of order coming in. My plan is to use exclusively stored procedures but to execute them dynamically using sp_executesql. I think I should use sp_executesql because that way I can have a config file (in xml) that I can store what stored procedures need to be called for which unit of work/order type/subsidiary. If I do this I should be able to easily configure each type of work to be done in a config file and let Service Broker handle the execution dynamically. As long as I keep the parameters the same for each of the stored procedures (I'm thinking maybe 4 or 5 parameters) and passing them to each of the stored procedures, this approach will allow me to dynamically configure Service Broker to do what it is supposed to do. I can pull what needs to be done out of the message that comes in with an XQuery expression on the config file. I know that I will have to configure my user (activation user) to be able to run sp_executesql and the security may be complex (especially since I'm using certificates). I can not use trusted databases. Are there any other considerations I should think about?
I have two databases Basket_ODS and Intelligence_ODS.
I created a user in the Basket_ODS and Intelligence_ODS databases as follows:
USE Basket_ods
GO
CREATE MASTER KEY ENCRYPTION BY PASSWORD = '*******'
CREATE USER BasketServiceUser WITHOUT LOGIN
ALTER AUTHORIZATION ON SERVICE::[Order Send] TO BasketServiceUser
GRANT CONTROL ON SERVICE::[Order Send]
TO BasketServiceUser
CREATE CERTIFICATE BasketServiceCertPriv
AUTHORIZATION BasketServiceUser
WITH SUBJECT = 'ForBasketService'
BACKUP CERTIFICATE BasketServiceCertPriv
TO FILE = 'BasketServiceCertPub'
In the other database...
I created the following:
USE Intelligence_ODS
GO
CREATE MASTER KEY ENCRYPTION BY PASSWORD = '************
USE Intelligence_ODS
GO
CREATE USER BasketServiceUser WITHOUT LOGIN
CREATE CERTIFICATE BasketServiceCertPub
AUTHORIZATION BasketServiceUser
FROM FILE = 'BasketServiceCertPub'
My Queue is in BASKET_ODS and is set up as:
ALTER QUEUE ODS.[Order Process Queue] WITH
ACTIVATION (
STATUS = ON,
PROCEDURE_NAME = ODS.ProcessOrderQueue,
MAX_QUEUE_READERS = 4,
EXECUTE AS 'BasketServiceUser'
)
I have performed the following grants in Basket_ODS
grant execute on ODS.ProcessOrderQueue to BasketServiceUser
ProcessOrderQueue calls [ODS].[MoveOrderTotals_Core] in the Intelligence_ODS database.
grant execute on [ODS].[MoveOrderTotals_Core] to BasketServiceUser
ProcessOrderQueue proc is set as follows:
ALTER procedure [ODS].[ProcessOrderQueue]
WITH EXECUTE AS 'BasketServiceUser'
[ODS].[MoveOrderTotals_Core] is set up as follows:
when I run ProcessOrderQueue I get an error message:
ALTER procedure [ODS].[MoveOrderTotals_Core](@Orderid uniqueidentifier)
with execute as 'BasketServiceUser'
I just don't understand when I run ProcessOrderQueue I get the following error message (when database trust is turned off)
The server principal "sa" is not able to access the database "Intelligence_ODS" under the current security context.
Can you help me figure out what I'm doing wrong. I've spent so much time on this security stuff. Is there another way to do this that is more straight forward without using database trust?
"Could not impersonate the client during assembly file operation."
The CLR function is invoked from Service Broker internal activation stored procedure.
"SELECT user_name()" returns dbo just before CREATE ASSEMBLY execution.
SqlContext.WindowsIdentity.Name is "NT AUTHORITYSYSTEM" as the Data Engine runs with the LocalSystem account.
How do I create a the necessary security context for "CREATE ASSEMBLY" to succeed ?
Service Broker Queue activation with EXECUTE AS = "SELF", "OWNER", domain account or dbo, all result in the above error. The Service Broker assembly having the internal activation stored procedure is registered "unsafe".
Okay, I have sort of a peculiar permissions question I am wondering if someone can help me with. I'm suspect there's a simple answer, but I'm unaware of it. Basically, here's the scenario...
I have a CLR stored procedure which does some dynamic SQL building based on values sent in via XML. It's a CLR stored procedure using XML because I want to build a parameterized statement (to guard against SQL Injection) based on a flexible number of parameters which are basically passed in the XML.
The dynamic SQL ends up reading from a table I'll call TableX and I actually discovered an (understandable) quirk with security.
Basically, the connection context is impersonating a low-privilaged Windows account ("UserX") coming from a .NET application. UserX has no permission to the table referenced in the dynamic SQL and because of the dyanmic nature of the query, the stored procedure apparently adopts the security context of UserX. Naturally, this throws a security exception saying UserX has no SELECT permission on TableX.
Now, I can give UserX read permission to the table in question to get things running, but one of the points of using stored procedures is to defer security to the procedure level vs. configuration for tables or columns.
So in striving toward my ideal of security at the procedure level, my question is what is the best way to allow minimum privilege in this case?
I thought about having the internals of the CLR stored procedure run under a different (low-privalaged) security context, but I am wondering if there's an alternate configuration that may use the same connection, and be as secure, but simpler.
I'm looking for some sample code rather than having to re-invent the wheel.
I need to write an analysis services stored proc that will invoke a SQL stored prod in my DW dataabse to retrieve a list of client_id's. I then need to construct and return a set object. This AS stored proc will be referenced from a role.
I've read few things about using the Set object rather than StrToSet function but then the only way I see to create a Member orSet object from a literal is to build an Expresion object and call the CalculateMdxObject(null).ToSet() function. Is this not equivalent to MDX.StrToSet()?
Also, ideally I'd like to connect to my SQL db by accessing the connection string from the Data Source objects in my AS DB. So far I have not found a way to do this.
Some guidance and sample code would be much appreciated.
I've been reseaching on SSB and have read quite a number of posts on this forum that closely relate to what im trying to achieve. I have a solution im designing that ideally consists of a central server (SQLENTERPRISE) that will receive messages asynchronously from remote clients (SQLEXPRESS) spanning a wide geographical region over a GPRS virtual private network on a TCP/IP transport. This ideally is a star and spoke architecture and requirements dictate high level security, no loss of messages whatsover as well as high reliability and scalability.
To meet the security requirement in the context of the above scenario, i was thinking implementing both dialog and endpoint security using certificates would be ideal. I've downloaded some samples and have encountered problems simulating the above scenario on 3 machines (I Server & 2 Clients). Client1 sends messags successfully, the 2nd client doesnt possibly because the certificate on the server matches that to client1 since i ran that script first. How do i make it that the Server shares one certificate with all remote clients? Or is there a better way to configure SSB to work in this scenario?
Secondly i have hard coded the Server's IP Address in the Routes created on the Clients. Considering this is over a GPRS Virtual network, how can i make the clients dyamically "discover" the Server? The idea here is to make adding and setting up of new remote clients easy so that you can just plug them in to the existing network
Any help or pointers would be greatly appreciated.
I have implemented an SSAS stored procedure for dynamic security and I call this stored procedure to obtain the allowed set filter. To my supprise, the stored procedure is being called repeatedly many times (more than 10) upon establishing the user session. Why is this happening?
I've created a custom dll using ADOMD.net that takes in a few parameters (Username being one of them) and returns a set of Allowed Members to apply dynamic dimension security. Example:
Code Snippet
Security.GetSetList(UserName,1,"SQL Server")
This works fine for users with less than, say, 100 tuples/members in their set....however for users with more an error:
Error ocurred retreiving child nodes: The 'S' attribute in the 'S' dimension has a generated dimension security expression that is not valid. ...
However, if I take the same set and explicitly (ie
Code Snippet
{[S].[1],[S].[2],.......[S].[50000]}) put it in the Allowed set there's no issue!
I noticed that it takes the UDF a couple of seconds longer to return for users with a large security context. Is there a time-out parameter that controls this? Is there a limitation in the amount of members I can add via custom dll (UDF)? (Some users can have up to 70,000 members)
I should also mention that the UDF uses the StrtoSet function to convert the security context string to a set. The string can be as large as 210,000 characters. Is there a limitation on the string size that can be passed in this function?
It seems simple but I'm not able to make it works. I've got a fact table with sites (DimImplantation) with allowed users (DimDroit) :
1 user can have access to many sites. I've tried these codes but only one worked and only in case where the user had only one site :
a) =DimImplantation[DimImplantation_ID]= LOOKUPVALUE (DimDroit[DimImplantation_ID]; DimDroit[Utilisateur];USERNAME() ) ==> worked for 1user with 1 site
b) =DimImplantation[DimImplantation_ID]= LOOKUPVALUE (DimDroit[DimImplantation_ID]; DimDroit[Utilisateur];USERNAME(); DimDroit[DimImplantation_ID];DimImplantation[DimImplantation_ID] ) ==> doesn't work and says 'not able to to identifie the value of DimImplantation[DimImplantation_ID] in the current context.
c) CALCULATE(VALUES(DimDroit[DimImplantation_ID], SUMMARIZE('DimDroit',DimDroit[Utilisateur]), DimDroit[Utilisateur]=USERNAME()) ==> version for SQL2012. I've had 'End of input reached'.
I have a business requirement to build a tabular data model, where I need to mask information of other Agents from a given Agent but I still need to show the overall sales of the given product.Â
For eg: IF an Agent is in APAC region he should see APAC region sales and also should be able see the sales of the same product in other region without knowing region specific break down.
For Agent  "Tom" in APAC region, the numbers will look like this APAC_Sales = 100,000 Other_Sales = 500,000
And if "John" is in NA region, then the number will look like this for him
NA_Sales    = 200,000 Other_Sales = 400,000
I wanted to create "Roles" based on the Region, so all the agents belong to "APAC" region will have same view as Tom and "NA" region agents will have John's view.
I have two different roles, each one with a dax filter. One is for filtering users that access by Excel, and other for filtering users that access by Reporting Services, respectively:
To create only one role that serves Excel and Reporting Services users, is it viable to use only the || (OR) operator?, is there any other regard i should take?
I was testing around with a sample service broker app using activation, and came across an interesting question. The little app sends a series of four messages to a queue, either on the same conversation or on seperate ones. Each message invokes one stored procedure in my activation procedure. All the procedure does is enter a record into a test table and then wait for an allotted amount of time. In my example, the first message called a proc that waited 20 sec, the 2nd one that waited 10 seconds, the third 5 seconds, and the 4th 1 second. I am using internal activation on the queue. It seemed that in both scenarios (sending on 4 separate conversations and on one conversation) the procedures executed "almost" sequentially. "Almost" meaning that the first procedure was done before the last one started executing. It makes sense to me that this would happen where I sent them on the same conversation, but not really when I sent them on 4 seperate ones. Is it because when I call a procedure from my activation procedure it locks the queue so that another message cannot be processed (I'm processing a message at a time)? How could I make it so that the 4th procedure (the one that only waits 1 second) returns before the 1st procedure (the one that waits 20 seconds)?
One is Basket_ODS and the other is Intelligence_ODS. I am using service broker activation on a queue to move data from the Basket_ODS table to the Intelligence_ODS database. Previously I was able to move from table to table in Basket_ODS, however now that I am moving it to another database on the same instance it is no longer working.
If I set my active connection in SQL Management Studio to this user(BrokerUser) and execute the "move" procedure it works. When activated by Service Broker however, it does not. Here is the error message:
2006-05-09 14:47:52.940 spid86s The activated proc [ODS].[ProcessOrderQueue] running on queue Basket_ODS.ODS.Order Process Queue output the following: 'The server principal "BrokerUser" is not able to access the database "Intelligence_ODS" under the current security context.'
I'm sure I missed something becasue it works fine in the same database. BrokerUser has datareader and datawriter in both databases.
We are trying to use xp_cmdshell commands in service broker. If I kick off the process without activation, the process succeeds. If I enable activation, the tasks with xp_cmdshell fail.
The xp_cmdshell task is either a echo command so that users know that we are processing a database, or a file copy.
I used the TechNet article by Roger Wolter to create the queues and stored procedures.
We will be working across domains that have one-way trusts.
I specify to execute as 'dbo' in the alter queue statement. I also define the activated sp to execute as 'dbo'.
But i keep getting permission errors from my activated sp. i have tried to excute as 'sa' , i have even tried to write a job that excutes to stored procedure but it also get weird errors. Bottom line if i exceute the sp in management studio logged in as sa it works , but thorugh activation or a job nothing works, as 'dbo' or 'sa'.
I need to change the Activation Key for a copy of SQL Server 2000 that I already have installed and up and running. I have the new key but can't figure out how to change from the old, no longer valid, key.
How can I change keys w/o dumping the DB, uninstalling, and reinstalling all of my instances?
I have seen the posting on determining how activation has failed and looking through the system logs is very helpful in determining why activation is not occuring, however, short of looking through the SQL Server logs is there another way to get the same information? Access to the SQL Server logs is fairly restricted. Does anyone know another way that this can be done? I have used the execute as technique described in the article on "Troubleshooting Activation Stored Procedures" and found it helpful in some cases.
I'm using service broker queue with internal activation to run a stored procedure. The DB server is windows 2003 R2, 4 cpu, with SQL server 2005 SP2. When I'm runing the stored procedure directly from the sql management studio it takes about 75% of the cpu and running for about a minute, but when the stored procedure is activated by the queue internal activation (as a background process) it uses only 25% of the machine cpu (my guess it uses only 1 cpu insted of all 4 cpu) and running for much longer time (sometimes even more than one hour). How can I change this behavior? I want it to run as fast as possible.
The queue decleration is:
CREATE QUEUE [TaskQueue] WITH ACTIVATION ( STATUS = ON, PROCEDURE_NAME = ProcessTasksProc, MAX_QUEUE_READERS = 1, EXECUTE AS SELF);
So is there something I'm supposed to configure to allow activation to fire after a cluster failover. We have had three where I have noticed that activation does not automatically fire back up. i have to alter the queue to get it going. None of the values on sys.service_queues imply that is disabled, but messages just keep stacking up into the queue and are not being dealt with. Once I alter the queue to turn activation on; everything takes off and starts working normaly.
Obviously, this is less than desireable for an environment that requires High availability.
Newbie question, but is there a way for me to set up a thread in my c# code so as it sleeps until it gets woken up by an event fired by an activation SP?
As in queue sits idle, thread sleeps. Queue receives a message, activation SP gets fired, activation SP raises an event which gets caught by event handler in code, which wakes up thread to do processing?
Currently I have a PC that has MSDE installed on it and is attached to database (MyData.MDF and a log file MYLog.LDF) located on its hard drive at c:data. When I detach from the database, place a copy of the two files noted above on my network drive @ u:data and try to attach I get the following error€™s:
SQL[1] exec error = -1: Changed database context to 'master'. €œDevice activation error. The physical file name u:dataMyData.MDF may be incorrect.€?
I have done some testing a have found that I can attach to a copy of my database if I move it anywhere on the c: drive, and or even to a 1Gb USB key attached to the system(e:). So far it seems to only be an issue if I move it to a mapped network drive. If anyone could please provide me with any info it would be greatly appreciated.
I'm having problems with activation. I have a CLR stored procedure that runs fine when run directly (and consumes messages from the queue). When I try to enable activation nothing happens. I've tried different execute as options and looked in the sql server logs and I don't see anything indicating why.
I've checked sys.service_queues and all options look correct including activation options. The sys.dm_broker_activated_tasks view is empty.
Without any errors I'm having difficulty tracking down the problem. I've seen references to service broker shutting down the procedure if it doesn't consume any messages but I haven't seen where this is indicated.
I have send a msg successful,but the proc failed. I altered the proc to correct the problem in the proc. The msg are still in the queue. Is there a manual way to excute the proc again to process the msg in the q? or do I wait for service broker to do it after a retry time (self recover)?
I am looking for an example of a SP that shows the best practices for internal activation? In BOL this topic describes the typical patter for reading messages from a queue. What is the typical pattern for reading messages from a queue using an internally activated SP? Do we still need to loop (considering the message arrival actually fired the sp)?
I have a set of service broker services setup that rely on external activation to process messages. I'm using the GotDotNet ExternalActivator, and it launches console applications that do the actual retrieval from the queues. The console applications are written to run continuously to avoid the cost of starting up .NET based console apps over and over again.
I am observing very odd timing behavior. With the receive queues empty and the external activator configured to run a minimum and maximum of 5 instances, I observe in SQL Profiler that most of the receive operations finish in about the same amount of time as my WAITFOR command in my receive stored procedure. However, there is usually one receive command that consistently takes upwards of 30 seconds and often causes sql timeout exceptions to be thrown. I know that I could code around this, but I wasn't really expecting this behavior.
Does anyone have any thoughts on why it might be occurring? I would have expected to routinely see my receive operations taking 15 seconds, give or take, especially when the queue is empty. Also, I have observed this behaviour on both SQL 2k5 Express and Dev Editions, so I don't think it's a version thing.
The stored procedure I am using to do the receive is:
BOL only seems to say that you can do it w/o really showing how, and the ExternalActivator sample at gotdotnet.com contains so much functionality I'm not sure what's required just for the external activation. Are there any docs or samples out there that focus on how to do it w/o obscuring the matter with a bunch of other functionality? (I prefer docs to project samples, b/c the samples tend to have hacks like hardcoded paths and connection strings so that they rarely work correctly right out of the box.)
OK, so assume I am recycling dialogs in my client code, and assume I am doing something similar to get a dialog handle in my TSQL. What should the activated stored procedure that is processing my queue look like if I am expecting thousands of messages per second? Assume also that there is a small bit of logic need to process each individual message? I am building for a high-throughput scenario and would like to get as much as possible out of each second-tier service broker server as possible before the aggregated data is moved up the chain to a master. The first tier is Express on a web server and exists primarily only as a forwarding mechanism.
I have an application that is set up using Service Broker to pass messages between services asynchronously. I am using event-based external activation and have successfully set up my event notification in SQL Server so if a message appears on any of my Service Broker queues, I'm getting the activation event from SQL Server sent to my activation service.
The problem that I am seeing is that every time I am posting a message onto a Service Broker queue, I am losing the event notification entry in the sys.event_notifications view and I'm not receiving my activation event notifications. When I execute the CREATE EVENT NOTIFICATION T-SQL statement to recreate the event notification, I'm getting the event notification immediately (since there are messages on the queues being monitored). The event notification appears to be registered until the next message is posted on the queue.
I am kinda curious how the rest of you are doing your error handling inside your activation stored procedures...best practices says you should not rollback the part of your transaction that receives the record off of the queue...but using a try...catch block will only allow you to rollback the entire transaction. I tried using savepoints and starting the try...catch after the savepoint and the proc still gives me the error: "
The current transaction cannot be committed and cannot be rolled back to a savepoint. Roll back the entire transaction." when the error scenario is encountered.
I tried using @@Error and checking for errors at the statement level which would allow me to do a partial rollback, but the type of errors I receive (i.e. invalid data types etc) are aborting the entire batch instead of passing the error and continuing.
Can other people on this forum using service broker give me an idea of how you are getting around this issue?
except for saying "Recieve Top(0)" which didn't make any sense to me. I have set the database to trustworthy (all of this is taking place within a single database on a local server). There are messages on the queue, and I have the queue activation set to max_queue_readers=2, procedure_name=StoredProcedure1, execute as owner. I tried execute as self and that didn't work either. I signed the assembly that contains StoredProcedure1 and the assembly that it references. The only thing that appears in the sql error log is this (I trimmed off the timestamp):
AppDomain 15 (TriggerTest.dbo[runtime].14) is marked for unload due to common language runtime (CLR) or security data definition language (DDL) operations. AppDomain 15 (TriggerTest.dbo[runtime].14) unloaded. AppDomain 18 (TriggerTest.dbo[runtime].17) created.
If I call the stored procedure manually it works just fine
