microsoft.public.sqlserver.notificationsvcs
[Top] [All Lists]

Re: NS Logic for Enabled Users

Subject: Re: NS Logic for Enabled Users
From: "Andy Wilbourn"
Date: Mon, 9 Jan 2006 12:55:50 -0500
Newsgroups: microsoft.public.sqlserver.notificationsvcs
Our fix, as stated before was to manually join back to the subscribers, 
through the views provided, to ensure they were enabled.

So is this a change to how NS is working for 2005, or possibly a bug? I may 
bring it up in the forum, since MSFT seems to respond more there. I really 
don't know where Joe Webb finds the time to monitor the newsgroups and 
forums. He is the one who does offer ideas.

"Vince Sefcik" <[email protected]> wrote in message 
news:eGN%[email protected]
> Some additional information on this issue:
>
> As I noted in my previous post, this issue (subscribers receiving 
> notifications even though their Enabled is False in the instance table 
> NSSubscribers) came up in a scheduled subscription.  This subscription 
> runs at 11:00:00 AM (UTC) every day.  Those subscribers with Enabled = 
> False that did receive notifications on 1/4/2006 at 11:00:00 AM did not 
> receive notifications on 1/5/2006 at 11:00:00 AM.  So, the notifications 
> that were sent when they shouldn't have been sent were sent only once. 
> Perhaps something was persisted in memory that got cleaned out after a 
> while?
>
> "Vince Sefcik" <[email protected]> wrote in message 
> news:[email protected]
>>I just ran into this same issue.  After converting a version 2.0 
>>application to SQL Server 2005, I imported a list of subscribers and 
>>subscriptions (using the 2005 API).  I set Enabled to False for all 
>>Subscribers except one (my account), yet three others received 
>>notifications.
>>
>> The instance table NSSubscribers column has a datetime setting of 
>> 1/3/2006 6:19 PM with Enabled = False for all Subscribers except me, yet 
>> the application table NS...Notifications has entires for those who 
>> received notifications and column SentTime has a datetime of 1/4/2006 
>> 11:01 AM. This is a scheduled notification.
>>
>> "Andy Wilbourn" <[email protected]> wrote in message 
>> news:[email protected]
>>>I have not had enough time to retest, we just changed our Match query to 
>>>join to the subscribers table and put the where enabled in our join.
>>>
>>> Looking at the view NS gives you it will not work. I am wondering if I 
>>> was just imagining that it worked before, but you seem to give me a 
>>> sanity check by your thinking it should work.
>>>
>>> "Joe Webb" <[email protected]> wrote in message 
>>> news:[email protected]
>>>> I'll see if I can repro - but I'm going to be out of town and largely
>>>> out of commission for the next 2 weeks or so. If you discover
>>>> something, would you post your findings?
>>>>
>>>> Thanks!
>>>>
>>>> -- 
>>>> Joe Webb
>>>> SQL Server MVP
>>>> http://www.sqlns.com
>>>>
>>>>
>>>> ~~~
>>>> Get up to speed quickly with SQLNS
>>>> http://www.amazon.com/exec/obidos/tg/detail/-/0972688811
>>>>
>>>> I support PASS, the Professional Association for SQL Server.
>>>> (www.sqlpass.org)
>>>>
>>>> On Mon, 5 Dec 2005 06:49:25 -0500, "Andy Wilbourn"
>>>> <[email protected]> wrote:
>>>>
>>>>>I looked at the view and it is only looking at my subscriptions table 
>>>>>where
>>>>>Enabled = 1. I will try to see if I can find what we did with prior 
>>>>>beta's
>>>>>to get it to work, or may just submit as a bug if warranted.
>>>>>
>>>>>Thanks for the help Joe.
>>>>>
>>>>>"Andy Wilbourn" <[email protected]> wrote in message
>>>>>news:OJVtu$Y%[email protected]
>>>>>> Yes I am. I will test it one more time in our QA environment, but our 
>>>>>> DEV
>>>>>> it is not working. We went through and did an update to the 
>>>>>> subscriber
>>>>>> table to set all disabled, we also tried using the NS API to do the 
>>>>>> same,
>>>>>> but our subscriptions were still coming back. I guess I can look at 
>>>>>> the
>>>>>> view in our application DB to see what the code is for ACTIVE
>>>>>> subscriptions.
>>>>>>
>>>>>> I am not sure if Shyam is monitoring this group anymore, he does not 
>>>>>> have
>>>>>> a primary focus of NS. Hopefully that does not get me in trouble, 
>>>>>> since I
>>>>>> don't know if that is common knowledge. We went to Redmond and fully
>>>>>> stress tested our solution for the SQL 2005, but did not test this 
>>>>>> aspect.
>>>>>> It came up when we were wanting to stress another part of our 
>>>>>> solution in
>>>>>> house, but did not want to generate alerts to all our test 
>>>>>> subscribers. To
>>>>>> our surprise we did generate millions of alerts.
>>>>>>
>>>>>>
>>>>>> "Joe Webb" <[email protected]> wrote in message
>>>>>> news:[email protected]
>>>>>>> Hi Andy -
>>>>>>>
>>>>>>> Your understanding of the Subscriber.Enabled property is correct. If
>>>>>>> the property is set to true, the subscriber's subscriptions are
>>>>>>> candidates for the matching rules and can generate notifications. If
>>>>>>> the property is set to false, the subscriptions associated with the
>>>>>>> disabled subscriber will not produce notifications.
>>>>>>>
>>>>>>> You are seeing something contrary to this in RTM?
>>>>>>>
>>>>>>> -- 
>>>>>>> Joe Webb
>>>>>>> SQL Server MVP
>>>>>>> http://www.sqlns.com
>>>>>>>
>>>>>>>
>>>>>>> ~~~
>>>>>>> Get up to speed quickly with SQLNS
>>>>>>> http://www.amazon.com/exec/obidos/tg/detail/-/0972688811
>>>>>>>
>>>>>>> I support PASS, the Professional Association for SQL Server.
>>>>>>> (www.sqlpass.org)
>>>>>>>
>>>>>>>
>>>>>>> On Fri, 2 Dec 2005 12:04:54 -0500, "Andy Wilbourn"
>>>>>>> <[email protected]> wrote:
>>>>>>>
>>>>>>>>I thought I had tested and validated that if a Subscriber is 
>>>>>>>>disabled,
>>>>>>>>then
>>>>>>>>none of their subscriptions would be matched without having to 
>>>>>>>>disable
>>>>>>>>the
>>>>>>>>subscription.
>>>>>>>>
>>>>>>>>I am using SQL 2005 NS. I believe we tested this on June CTP and it 
>>>>>>>>was
>>>>>>>>working, but now with RTM it is not.
>>>>>>>>
>>>>>>>>Can someone tell what is supposed to happen?
>>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>
>>>
>>
>>
>
> 



<Prev in Thread] Current Thread [Next in Thread>